Image header of WCCI 2016
Background image

Welcome to IEEE WCCI 2016 Programs

Background image

Program-at-a-Glance

Background image

Technical Programs

Background image

Technical Programs

Background image

Social Events

Background image

Workshops

Background image

Tutorials

Background image

Special Sessions

Background image

Panel Sessions

Background image

Competitions

FIND US ON

line line

General Information

Special session proposals are invited to the 2016 IEEE WCCI. Proposals should include the title, aim and scope, list of main topics, and the names and short biography of the organizers.

All special sessions proposals should be submitted to the following Special Sessions Chairs before 15 November 2015.

Dr. Zhi-Hua Zhou (For neural networks and learning systems related topics, hybrid of neural networks, learning systems and computational intelligence technologies). Papers submitted to this special session track (if accepted and presented) will be published in the IJCNN proceedings.

IJCNN 2016 Special Sessions

  • IJCNN-01 Mind, Brain, and Cognitive Algorithms
  • IJCNN-02 Nature-Inspired Neural Network Optimization
  • IJCNN-03 Machine Learning Methods Applied to Vision and Robotics (MLMVR)
  • IJCNN-04 Machine Learning for Computer Vision
  • IJCNN-05 Optimizing Neural Networks via Evolutionary Computation and Swarm Intelligence
  • IJCNN-06 Computational Intelligence for Big Social Data Analysis
  • IJCNN-07 Probabilistic Models and Kernel Methods
  • IJCNN-08 Advanced Learning for Large-Scale Heterogeneous Computing
  • IJCNN-09 Intelligent Vehicle and Transportation Systems
  • IJCNN-10 Visual Semantic Learning from Big Surveillance Data
  • IJCNN-11 Human-like Learning for Intelligent Systems
  • IJCNN-12 Deep Learning for Brain-Like Computing and Pattern Recognition
  • IJCNN-13 Advances in Computational Intelligence for Applied Time Series Forecasting (ACIATSF)
  • IJCNN-14 Advanced Supervised Learning Techniques and Its Applications
  • IJCNN-15 Machine Learning with Incompletely Labeled Data
  • IJCNN-16 Applications of Machine Learning Techniques to Bioinformatics
  • IJCNN-17 Complex-Valued Neural Networks
  • IJCNN-18 Neural Networks: Model and Application in Real World Problems
  • IJCNN-19 Ordinal Regression and Ranking
  • IJCNN-20 Computational Intelligence Algorithms for Digital Audio Applications
  • IJCNN-21 Deep Learning, Medical Imaging, and Translational Medicine
  • IJCNN-22 Sequential Learning with Neural Networks
  • IJCNN-23 Research on Large Scale Clustering Algorithms
  • IJCNN-24 Deep Learning for Big Multimedia Understanding
  • IJCNN-25 Distributed Learning Algorithms for Neural Networks
  • IJCNN-26 Concept Drift, Domain Adaptaion and Learning in Dynamic Environments
  • IJCNN-27 Extreme Learning Machines (ELM)
  • IJCNN-28 Biologically-inspired Neural Networks for Robotics
  • IJCNN-29 Reinforcement Learning and Approximate Dynamic Programming for Optimization in Dynamic Environment
  • IJCNN-30 Biologically Inspired Computational Vision
  • IJCNN-31 Privacy-preserving in Data Mining and Applications
  • IJCNN-32 Advanced Machine Learning Methods and Applications from Complicated Data Environment
  • IJCNN-33 Computational Intelligence and Machine Learning for Power Grid Security
  • IJCNN-34 Smart Educational Techniques in Big Data Age
  • IJCNN-35 Spiking Neural Networks
  • IJCNN-36 Online Real-Time Learning Strategies for Large Data Streams
  • IJCNN-37 Approximate/Adaptive Dynamic Programming for Control of Cyber-physical Systems
  • IJCNN-38 Energy Efficient Deep Neural Networks
  • IJCNN-39 Hybrid Neural Intelligent Systems
  • IJCNN-40 Computational Intelligence for Safer and Smarter Societies
  • IJCNN-41 Incremental Machine Learning: Methods and Applications
  • IJCNN-42 Theoretical Foundations of Deep Learning Models and Algorithms
  • IJCNN-43 Neural Network Transfer Learning for the Recognition of Human Behavior and Affect
  • IJCNN-44 Deep Reinforcement Learning (DRL)
  • IJCNN-45 Neuro-Inspired Computing with Beyond-CMOS Technology
  • IJCNN-46 Dynamics and Design of Neural Networks and Applications in Industry
  • IJCNN-47 Computational Intelligence Paradigms For Space Weather Prediction
  • IJCNN-48 Affective Brain-Computer Interaction
  • IJCNN-49 Machine Learning Methods Robust to Large Outliers
  • IJCNN-50 Computational Intelligence for Personal Health
  • IJCNN-51 Advanced Methods in Optimization and Machine Learning for Multimedia Computing
  • IJCNN-52 Approximate Dynamic Programming and Reinforcement Learning for High Dimensional Systems

Dr. Uzay Kaymak (For fuzzy systems related topics, hybrid of fuzzy systems and computational intelligence technologies). Papers submitted to this special session track (if accepted and presented) will be published in the FUZZ-IEEE proceedings.

  • FUZZ-IEEE-01 Uncertainty Theory and Its Application
  • FUZZ-IEEE-02 Belief Function Theory and Its Applications
  • FUZZ-IEEE-03 Recent Trends in Many-Valued Logic and Fuzziness
  • FUZZ-IEEE-04 Fuzzy Interpolation
  • FUZZ-IEEE-05 Inter-Relation Between Interval and Fuzzy Techniques
  • FUZZ-IEEE-06 Software for Soft Computing
  • FUZZ-IEEE-07 Recent Advances and New Challenges in Evolving Fuzzy Systems
  • FUZZ-IEEE-08 Advances to Type-2 Fuzzy Logic Control
  • FUZZ-IEEE-09 Simulation Modeling and Fuzzy Logic
  • FUZZ-IEEE-10 Type-2 Fuzzy Sets and Systems Applications (T2-A)
  • FUZZ-IEEE-11 Fuzzy and Intelligent Control Systems
  • FUZZ-IEEE-12 The Theory of Type-2 Sets and Systems (T2-T)
  • FUZZ-IEEE-13 Fuzzy Set Theory in Computer Vision
  • FUZZ-IEEE-14 Fuzzy Systems on Renewable Energy
  • FUZZ-IEEE-15 Methods and Applications of Fuzzy Cognitive Maps
  • FUZZ-IEEE-16 Evolutionary Fuzzy Systems
  • FUZZ-IEEE-17 Linguistic Summarization and Description of Data
  • FUZZ-IEEE-18 Handling Uncertainties in Big Data by Fuzzy Systems
  • FUZZ-IEEE-19 Interval-Valued Fuzzy Sets: Theory and Applications
  • FUZZ-IEEE-20 Intelligent Medical Science
  • FUZZ-IEEE-21 Adaptive Fuzzy Control for Nonlinear Systems
  • FUZZ-IEEE-22 Fuzzy Decision-Making: Consensus and Missing Preferences
  • FUZZ-IEEE-23 Bio-inspired Fuzzy Logic Approaches - Interdisciplinary Emergent Technologies
  • FUZZ-IEEE-24 Medical Image Analysis based Computational Intelligence Techniques
  • FUZZ-IEEE-25 Complex Fuzzy Sets and Logic
  • FUZZ-IEEE-26 From Type-1 to Type-n Fuzzy Systems Modeling
  • FUZZ-IEEE-27 Fuzzy Logic and Computational Intelligence Applications in Construction Engineering and Management
  • FUZZ-IEEE-28 Fuzzy-based Methods for Machine Learning: Data Preprocessing, Learning Models and Their Applications
  • FUZZ-IEEE-29 Recent Advances in Fuzzy Control System Design and Analysis
  • FUZZ-IEEE-30 Fuzzy Approaches for Advanced Manufacturing
  • FUZZ-IEEE-31 New Approaches to Fuzzy Web Intelligence
  • FUZZ-IEEE-32 Fuzzy Pattern Recognition for Big Data Modeling and Data Mining
  • FUZZ-IEEE-33 Information Fusion and Fuzzy Linguistic Decision Making
  • FUZZ-IEEE-34 Methods of Operations Research For Decision Support Under Uncertainty
  • FUZZ-IEEE-35 Moving Towards Neutrosophic Logic
  • FUZZ-IEEE-36 Fuzzy Social Network Analysis

Dr. Mengjie Zhang (For evolutionary computation related topics, hybrid of evolutionary computation and computational intelligence technologies). Papers submitted to this special session track (if accepted and presented) will be published in the IEEE CEC proceedings.

IEEE CEC 2016 Special Sessions

  • CEC-01 Optimization Methods in Energy Internet System
  • CEC-02 Intelligent Evaluation of Complex Algorithms (IEOCA)
  • CEC-03 Nature-Inspired Constrained Single- and Many-Objective Optimization
  • CEC-04 Evolutionary Computer Vision
  • CEC-05 Evolutionary Scheduling and Combinatorial Optimization
  • CEC-06 Evolutionary Feature Selection and Construction
  • CEC-07 Evolutionary Bilevel Optimization
  • CEC-08 Many-Objective Optimization
  • CEC-09 Evolutionary Computation for Nonlinear Equation Systems
  • CEC-10 Computational Intelligence in Aerospace Science and Engineering
  • CEC-11 Automated Design: Hyper-heuristics and Metaheuristics
  • CEC-12 Brain Storm Optimization Algorithms
  • CEC-13 Evolutionary Computation and Big Data
  • CEC-14 Fireworks Algorithm for Its Application on Big-Data
  • CEC-15 Evolutionary Optimization for Non-Convex Machine Learning
  • CEC-16 Fitness Landscape Analysis in Practice
  • CEC-17 Computationally Expensive Optimization
  • CEC-18 Big Optimization (BigOpt2016)
  • CEC-19 Efficient Non-dominated Sorting and Pareto Approaches to Many-Objective Optimization
  • CEC-20 When Evolutionary Computation Meets Data Mining
  • CEC-21 Evolutionary Computation for the Design of Digital Filters
  • CEC-22 Evolutionary Robotics
  • CEC-23 Memetic Computing
  • CEC-24 Evolutionary Computation in Operations Research, Management Science and Decision Making
  • CEC-25 Bat Algorithm and Cuckoo Search
  • CEC-26 Evolutionary Computation and Other Computational Intelligence Techniques for Cyber Security
  • CEC-27 Evolutionary Computation in Dynamic and Uncertain Environments
  • CEC-28 Complex Networks and Evolutionary Computation
  • CEC-29 Evolutionary Computation for Computational Biology
  • CEC-30 Optimization, Learning, and Decision-Making in Bioinformatics and Bioengineering (OLDBB)
  • CEC-31 New Directions in Evolutionary Machine Learning
  • CEC-32 Differential Evolution: Past, Present and Future
  • CEC-33 Evolutionary Algorithms for Mixed-Integer Optimization Problems
  • CEC-34 Evolutionary Multi-objective Optimization
  • CEC-35 Parallel and Distributed Evolutionary Computation in the Inter-Cloud Era
  • CEC-36 Evolutionary Computation in Architectural Design
  • CEC-37 Theoretical Foundations of Bio-inspired Computation
  • CEC-38 Evolutionary computation with human factors
  • CEC-39 Transfer Learning in Evolutionary Computation
  • CEC-40 Advances in Decomposition­based Evolutionary Multi­objective Optimization (ADEMO)
  • CEC-41 Niching Methods for Multimodal Optimization
  • CEC-42 Special Session Associated with Competition on Bound Constrained Single Objective Numerical Optimization
  • CEC-43 Evolutionary Physical Systems and Matter
  • CEC-44 Large Scale Global Optimization
  • CEC-45 Evolutionary Computing Application in Hardware
  • CEC-46 Bio-inspired and Evolutionary Computation methods for unsupervised learning
  • CEC-47 Quantum Computing and Evolutionary Computation
  • CEC-48 Hybrid Cultural Algorithms: Beyond Classical Cultural Algorithms
  • CEC-49 Genetic Improvement of Software + Search-Based Software Engineering (GI+SBSE)
  • CEC-50 Dynamic Multi-objective Optimization
  • CEC-51 Evolutionary Computation for Human Center Decision Making Systems: Trends and Applications
  • CEC-52 Hidden Complex Networks in Evolutionary Dynamics. Past, Present and Future.
  • CEC-53 Multi- Fidelity Design Optimization under Epistemic Uncertainties
  • CEC-54 Geometrical and Topological methods in Evolutionary Computing
  • CEC-55 Multiobjective Optimization with Surrogate Models

Dr. Chuan-Kang Ting (For cross-disciplinary and computational intelligence applications). Papers submitted to this cross-disciplinary and CI applications special session track (if accepted and presented) will be published in one of the three conference proceedings (IJCNN, Fuzzy-IEEE, or IEEE CEC) that is most appropriate to the papers. Such decision will be made by the Special Session Organizers in consultation with the Special Session Chair and one of the three Conference Chairs.

IEEE WCCi 2016 Cross-Disciplinary and Computational Intelligence Applications Special Sessions

  • CDCI-01 Interactive Computational Intelligence for Industry 4.0
  • CDCI-02 Computational Intelligence for Physiological and Affective Computing (CIPAC)
  • CDCI-03 Computational Intelligence methods for Natural Language Processing
  • CDCI-04 Computational Intelligence in Bioinformatics (CIB)
  • CDCI-05 Computational Intelligence and Games
  • CDCI-06 Computational Intelligence for Music, Art, and Creativity
  • CDCI-07 Advanced Computational Intelligence Methods for Health Technologies and Applications
  • CDCI-08 Computational Intelligence Techniques for the Analysis of Big and Streaming Data in Complex Systems
  • CDCI-09 Computational Intelligence for Economics and Finance
  • CDCI-10 Computational Intelligence Methodologies for Environmental Sustainability and Sustainable Development: Theory and Applications
  • CDCI-11 Computational Intelligence in Ecological Informatics and Environmental Modelling
  • CDCI-12 From Big Data to Big Knowledge Using Computational Intelligence in Biomedicine
  • CDCI-13 Computational Intelligence in Power System
  • CDCI-14 Computational Intelligence for Security, Surveillance, and Defense
  • CDCI-15 Computational Intelligence algorithms for industrial and design Engineering Problems
  • CDCI-16 Computational Intelligence in Dynamics of ComplexNetworks: Models and Applications
  • CDCI-17 Computational Intelligence Methods Accelerated on Parallel and Distributed Architectures for Applications in Bioinformatics, Computational Biology and Systems Biology
  • CDCI-18 Computational Intelligence for Unmanned Systems
  • CDCI-19 Computational Intelligence in Biometrics and Biometrics Applications
  • CDCI-20 Computational Intelligence Algorithms in Wireless Sensor Networks
  • CDCI-21 Computational Intelligence for Pathology Informatics
  • CDCI-22 Intelligent Control of Unmanned Surface Vehicles
  • CDCI-23 Computational Intelligence for Knowledge and Skills Transfer: Theories, Algorithms and Applications
  • CDCI-24 Cognitive Agent and Robotics for Human-Centric Systems: New Models and Challenges
  • CDCI-25 Computationally Intelligent Methods in Neural Information Processing
  • CDCI-26 Cognitive Robotics
  • CDCI-27 Deep Computational Intelligence Models
  • CDCI-28 Applied Computational Intelligence for the Economics of Financial Market Infrastructures
  • CDCI-29 Computational Intelligence in Marketing and Social Sciences
  • CDCI-30 Human Symbiotic Systems
  • CDCI-31 Computational Intelligence, Nature-Inspired Learning and Big Data
  • CDCI-32 Robust Data Mining Techniques through Hybrid Metaheuristic (RDMM)

FIND US ON

line line

IJCNN Special Sessions


IJCNN 2016 Special Sessions

IJCNN-01 Mind, Brain, and Cognitive Algorithms

Organized by Leonid Perlovsky, José F. Fontanari, Asim Roy, Angelo Cangelosi and Daniel Levine

Recent progress opens new directions for modeling the mind and brain and developing cognitive algorithms for engineering applications. Cognitive algorithms solve traditional engineering problems much better than before, and new areas of engineering are opened modeling human abilities in cognition, emotion, language, art, music, cultures. Cognitive dissonances and behavioral economics is another new active area of research. A wealth of data are available about the ways humans perform various cognitive tasks (e.g., scene and object recognition, language acquisition, interaction of cognition and language, music cognition, cognitive dissonance) as well as about the biases involved in human judgment and decision making (e.g., the prospect theory and the fuzzy-trace theory). A wealth of data on the web can be exploited for extracting cognitive data. Explaining these laws and biases using realistic neural networks architectures, including neural modeling fields, as well as more traditional learning algorithms requires a multidisciplinary effort.

Scope and Topics

The aim of this special session is to provide a forum for the presentation of the latest data, results, and future research directions on the mathematical modeling of higher cognitive functions using neural networks, neural modeling fields, as well as cognitive algorithms exploiting web data and solving traditional and new emerging engineering problems, including genetic association studies, medical applications, Deep Learning, and Big Data. The special session invites submissions in any of the following areas:

  • Neural network models of higher cognitive function
  • Neural mechanisms of emotions, cognition
  • Embodied cognition modeling
  • Neural modeling fields (NMF)
  • Perceptual processing
  • Language learning
  • Cognitive and emotional processing
  • Cognitive models of decision-making
  • Models of emotional mechanisms
  • Models of cognitive dissonances
  • Cognitive, language, and emotional models of cultures
  • Cognitive functions of art, music, and spiritual emotions.
  • Emotions in cognition (affective cognition)
  • Cognitive dissonance, neural models
  • Cognition and cultures
  • Medical applications
  • Genome association studies
  • Big Data

Keywords: Cognition, Emotions, Decision-Making, Dynamic Logic, Language Acquisition, Language Emotionality, Cognitive Dissonance, Music Cognition, Models of Cultures, Neural Modeling Fields, ART Neural Network, Fuzzy-Trace Theory, Prospect Theory, Deep Learning, Genome Associations, Big Data.

IJCNN-02 Nature-Inspired Neural Network Optimization

Organized by Anna Rakitianskaia and Andries Engelbrecht

Nature-inspired algorithms have been successfully applied to neural network training, neural network architecture optimization, and neural network architecture construction in the past. Applications of nature- inspired algorithms to neural networks are diverse and often hybridized with more traditional gradient-descent based methods . Compared to gradient-based methods, nature-inspired algorithms are less sensitive to weight initialisation, less likely to become trapped in local optima, and independent of the activation function gradient. Despite the relative success of nature-inspired algorithms in the neural network context, a solid theoretical foundation for such applications is often lacking. Successful applications of nature-inspired methods to newer neural network paradigms such as deep learning are yet to be seen. Some nature-inspired algorithms were shown to suffer from stagnation when applied to neural networks. Optimizing large real-world neural networks is a challenging task due to the inherent high dimensionality of the weight space, high correlation between the individual weights, and our limited knowledge of the error landscape properties in high dimensions. The proposed nature-inspired methods must scale well to high dimensions to be usable in a real-world context.

Scope and Topics

The aim of this special session is to investigate the existing nature-inspired approaches to neural network optimization, to develop new efficient nature-inspired neural network training and architecture optimization algorithms, to encourage discussion of the existing challenges, to identify problems, and to propose solutions. The proposed special session will provide an excellent forum for fellow researchers in this exciting cross-disciplinary field. The topics of the special session include, but are not limited to:

  • Swarm-based algorithms (PSO, ACO, etc.) for neural network optimization
  • Evolutionary algorithms (EA, ES, DE, etc.) for neural network optimization
  • Hybrid nature-inspired algorithms for neural network optimization
  • Scalability analysis of nature-inspired methods in the neural network context
  • Applications of nature-inspired algorithms to deep learning
  • Real-world neural networks applications of nature-inspired algorithms
  • Theoretical analysis of nature-inspired algorithms in the neural network context
  • Hyper-heuristics for neural network training.
  • Neural network training and architecture optimization in dynamic environments
  • Empirical analysis of nature-inspired neural network algorithms

IJCNN-03 Machine Learning Methods Applied to Vision and Robotics (MLMVR)

Organized by José García-Rodríguez, Sergio Escalera, Alexandra Psarrou, Isabelle Guyon and Andrew Lewis

Over the last decades there has been an increasing interest in using machine learning methods combined with computer vision techniques to create autonomous systems that solve vision problems in different fields. This special session is designed to serve researchers and developers to publish original, innovative and state-of-the art algorithms and architectures for real time applications in the areas of computer vision, image processing, biometrics, virtual and augmented reality, neural networks, intelligent interfaces and biomimetic object-vision recognition.

This special session provides a platform for academics, developers, and industry- related researchers belonging to the vast communities of *Neural Networks*, *Computational Intelligence*, *Machine Learning*, *Biometrics*, *Vision systems*, and *Robotics *, to discuss, share experience and explore traditional and new areas of the computer vision and machine learning combined to solve a range of problems. The objective of the workshop is to integrate the growing international community of researchers working on the application of Machine Learning Methods in Vision and Robotics to a fruitful discussion on the evolution and the benefits of this technology to the society.

Scope and Topics

The Special Session topics can be identified by, but are not limited to, the following subjects:

  • Artificial Vision
  • Video tracking
  • 3D Scene reconstruction
  • 3D Tracking in Virtual Reality Environments
  • 3D Volume visualization
  • Computational Intelligence
  • Machine Learning
  • Intelligent Interfaces (User-friendly Man Machine Interface)
  • Self-adaptation and self-organisational systems
  • Multi-camera and RGB-D camera systems
  • Robust computer vision algorithms (operation under variable conditions,
  • object tracking, behaviour analysis and learning, scene segmentation)
  • Multi-modal Human Pose Recovery and Behavior Analysis
  • Gesture and posture analysis and recognition
  • Biometric Identification and Recognition
  • Extraction of Biometric Features (fingerprint, iris, face, voice, palm, gait)
  • Surveillance systems
  • Robotics vision
  • Hardware implementation and algorithms acceleration (GPUs, FPGA,s,...)

IJCNN-04 Machine Learning for Computer Vision

Organized by Brijesh Verma, and Mohammed Bennamoun

There is a great interest of machine learning algorithms among the computer vision researchers. Many machine learning algorithms have successfully demonstrated the capability of solving real world problems in computer vision field. The purpose of the special session on Machine Learning for Computer Vision is to address the latest developments of machine learning algorithms for numerous applications in the computer vision.

Scope and Topics

This session aims to bring together machine learning and computer vision researchers to demonstrate latest progress, emphasize new research questions and collaborate for promising future research direction.
The theme of the session is the application of machine learning to computer vision. The list of topics includes and is not restricted to the following:

  • Neural learning
  • deep learning
  • feature learning
  • metric learning
  • representation learning
  • probabilistic graphical models
  • manifold learning
  • hybrid learning
  • supervised learning
  • unsupervised learning
  • support vector machines
  • statistical learning and reinforcement learning for image processing
  • document processing
  • pattern recognition
  • audio/video processing
  • medical imaging
  • object recognition
  • face recognition
  • image analysis

IJCNN-05 Optimizing Neural Networks via Evolutionary Computation and Swarm Intelligence

Organized by Wei-Chang Yeh, Liang Feng and Yew-Soon Ong

Today, neural networks have been widely recognized as useful frameworks to model multidimensional nonlinear relationships. It has been successfully applied in real-world applications including signal processing, robot control, classification, etc. Recently, it has also been employed to construct deep architectures for deep learning to model high-level abstractions in data, and achieved considerable success in applications such as natural language processing, music signal recognition, computer vision and automatic speech recognition, etc. Despite the success achieved by neural network, constructing multilayer neural networks involves challenging optimization problems, i.e., finding appropriate architecture and the corresponding optimal weights for some of the core applications of interest.

Evolutionary Computation and Swarm Intelligence are natural inspired heuristic methods with global search capability that have attracted extensive attentions in the last decades. They have been successfully applied to complex optimization problems including continuous optimization, combinatorial optimization, constrained optimization, etc. The aim of this special session is to provide a forum for researchers in the field of neural network to exchange their latest advances in theories, technologies, and practice of optimizing neural networks, especially with deep and large architecture, using evolutionary computation and swarm intelligence.

Scope and Topics

This Special Session on “Optimizing Neural Networks via Evolutionary Computation and Swarm Intelligence” mainly focus on the research of exploring Evolutionary Computation and Swarm Intelligence methodologies for optimizing the neural network architectures. Despite a significant amount of research have been done in neural networks, there remains many open issues and intriguing challenges in optimizing neural network architectures, especially in todays’ deep learning context, where neural networks usually have many layers and large number of neurons.
Authors are invited to submit their original and unpublished work in the areas including, but not limited to:

  • Evolutionary Computation in Neural Networks,
  • Swarm Intelligence in Neural Networks,
  • Advances in Evolutionary Computation and/or Swarm Intelligence,
  • Knowledge incorporation in Evolutionary Computation and/or Swarm Intelligence,
  • Advances in Neural Networks
  • Analytical studies that enhance our understanding on the behaviors of Evolutionary Computation and/or Swarm Intelligence in optimizing Neural Networks,
  • Novel or Improved frameworks of Neural Networks,
  • Others.

IJCNN-06 Computational Intelligence for Big Social Data Analysis

Organized by Erik Cambria, Amir Hussain and Newton Howard

As the Web rapidly evolves, Web users are evolving with it. In an era of social connectedness, people are becoming increasingly enthusiastic about interacting, sharing, and collaborating through social networks, online communities, blogs, Wikis, and other online collaborative media. In recent years, this collective intelligence has spread to many different areas, with particular focus on fields related to everyday life such as commerce, tourism, education, and health, causing the size of the Social Web to expand exponentially.

The distillation of knowledge from such a large amount of unstructured information, however, is an extremely difficult task, as the contents of today’s Web are perfectly suitable for human consumption, but remain hardly accessible to machines. The opportunity to capture the opinions of the general public about social events, political movements, company strategies, marketing campaigns, and product preferences has raised growing interest both within the scientific community, leading to many exciting open challenges, as well as in the business world, due to the remarkable benefits to be had from marketing and financial market prediction.

The main aim of this Special Session is to explore the new frontiers of big data computing for opinion mining and sentiment analysis through computational intelligence techniques, in order to more efficiently retrieve and extract social information from the Web.

Scope and Topics

The Special Session aims to provide an international forum for researchers in the field of big data computing for opinion mining and sentiment analysis to share information on their latest investigations in social information retrieval and their applications both in academic research areas and industrial sectors. The broader context of the Special Session comprehends information retrieval, natural language processing, web mining, semantic web, and computational intelligence. Topics of interest include but are not limited to:

  • Computational intelligence for sentiment mining
  • Concept-level sentiment analysis
  • Biologically-inspired opinion mining
  • Sentiment identification & classification
  • Association rule learning for opinion mining
  • Time evolving opinion & sentiment analysis
  • Multi-modal sentiment analysis
  • Multi-domain & cross-domain evaluation
  • Knowledge base construction & integration with opinion analysis
  • Transfer learning of opinion & sentiment with knowledge bases
  • Sentiment topic detection & trend discovery
  • Social ranking
  • Social network analysis
  • Opinion spam detection

IJCNN-07 Probabilistic Models and Kernel Methods

Organized by Shiliang Sun, Yuanbin Wu, Huawen Liu and Yong Ma

Probabilistic models and kernel methods are two of the core techniques in machine learning and pattern recognition. During the past twenty years, many successful applications based on them have been developed. Probabilistic models are known as a solid framework to model uncertainty and dependency, while kernel methods provide a powerful approach to modeling nonlinear relationship through the use of linear methods and the kernel trick. These two types of techniques are closely related and some methods can be understood from both sides. Building good and scalable models and giving effective and efficient inference methods are important concerns for research on probabilistic models; creatively applying kernel methods to solve problems such as those in semi-supervised learning, multi-view learning, transfer learning, and multi-task learning is also an active research field. This special session intends to provide a platform for researchers to discuss and report related progresses.

Scope and Topics

The special session covers theory, models, algorithms and applications for probabilistic models and kernel methods. Typical topics include the following (but not limited to):

  • Gaussian processes
  • Hidden Markov models
  • Conditional random fields
  • Topic models
  • Deep models
  • Maximum entropy discrimination
  • Support vector machines
  • Nyström methods
  • Structured prediction
  • Statistical learning theory
  • Multi-view learning
  • Manifold learning
  • Multi-task learning
  • Transfer learning
  • Active learning
  • Semi-supervised learning
  • EEG-based brain-computer interfaces
  • Human activity analysis from images and videos
  • Natural language processing
  • Education data analysis

IJCNN-08 Advanced Learning for Large-Scale Heterogeneous Computing

Organized by Quan Zou

With the age of big data upon us, machine learning techniques have met tremendous challenges. Data comes from multiple sources and becomes large and hybrid. Due to the diversity of data acquisition and storage, heterogeneous information is widely used for recording and conveying semantic nowadays. It appears dramatically on multi-media and bioinformatics data. Advanced machine learning techniques have developed quickly in recent years. Several impacted new methods were reported in the high-level journals and conferences. For example, affinity propagation was published in Science as a novel clustering algorithm. Extreme learning machine was proposed in Neurocomputing, and became the most cited and downloaded paper. Recently, deep learning has become to be the hot topic and seems to be suitable for big multi-media data. Parallel mechanism is also developed by the scholar and industry researchers, such as Mahout. More and more computer scientists devoted to the advanced large-scale and heterogeneous machine learning techniques. However, application in real world fell behind the technique growth.

This special session will target the recent large scale machine learning techniques together with application. We especially welcome novel classification and clustering algorithms, such as learning strategies for large-scale hybrid and heterogeneous data, strategies for large-scale imbalanced learning, strategies for multiple views learning, strategies for various semi-supervised learning, strategies for multiple kernels learning, etc. Application on multi-media and biology scalable hybrid data is encouraged. Only machine learning theory without real world application cannot be accepted. We also encourage authors to supply their codes and open their real data, which would make our issue more impacted. Please do not test your algorithm just only on UCI or benchmark data.

Scope and Topics

The editors expect to collect a set of recent advances in the related topics, to provide a platform for researchers to exchange their innovative ideas and real application data. Typical topics include the following (but not limited to):

  • Large scale classification algorithms with application
  • Large scale clustering algorithms with application
  • Deep learning strategies for hybrid and heterogeneous data
  • Imbalanced learning algorithms for bioinformatics data
  • Multiple views leaning for image classification
  • Semi supervised learning strategies for big hybrid data
  • Ensemble learning strategies for big hybrid data
  • Parallel learning techniques for ultra large data
  • Multiple kernels learning with application
  • Multiple labels classification algorithms with application
  • Extreme learning techniques and application on multi-media or bio-big data

IJCNN-09 Intelligent Vehicle and Transportation Systems

Organized by Yi Lu Murphey, Mahmoud Abou-Nasr, Ishwar K Sethi, Robert Karlsen, Ana Bazzan and Chaomin Luo

The research and development of intelligent vehicles and transportation systems are rapidly growing worldwide. Intelligent transportation systems are making transformative changes in all aspects of surface transportation based on vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) connectivity. With the decreasing sensor costs and computer chips, and increasing computing power and data storage capacity, it has become practical to build a host of intelligent devices in cars that can be used in airbag control, unwelcome intrusion detection, collision warning and avoidance, power management and navigation, driver alertness monitoring etc. Computational intelligence plays a vital role in building all types and levels of intelligence in vehicle and transportation systems.

The objective of this special session is to provide a forum for researchers and practitioners to present advanced research in computational intelligence with a focus on innovative applications to intelligent vehicle and transportation systems.

Scope and Topics

This session seeks contribution on the latest developments and emerging research in all aspects of intelligent vehicle and transportation systems. Specific topics for the session include, but are not limited to:

  • Advanced transportation information and communication systems
  • Cloud computing and big data in transportation and vehicle systems
  • Multimodal intelligent transport systems and services
  • Personalized driver and traveler support systems
  • Pervasive and ubiquitous computing in logistics
  • Simulation and forecasting models
  • Spatio-temporal traffic pattern recognition
  • Connected vehicles of the future
  • Air, Road, and Rail Traffic Management
  • Advanced Transportation Management
  • Collision detection and avoidance
  • vehicle communications and connectivity
  • Driver state detection and monitoring
  • Driver assistance and automation systems
  • Vehicle fault diagnostics and health monitoring
  • Automated driving and driverless car
  • Learning and adaptive Control
  • Object recognitions such as pedestrian detection, traffic sign detection and recognition
  • Route guidance systems
  • Trip modeling and driver speed prediction
  • Vehicle energy management and optimization in hybrid vehicles

IJCNN-10 Visual Semantic Learning from Big Surveillance Data

Organized by Zhaoxiang Zhang, Xiang Bai, Rongrong Ji and Chuanping Hu

Intelligent video surveillance is an important topic in the field of computer vision and pattern recognition. Significant progress has achieved in the last decades, from object detection, tracking and parsing to activity recognition and video understanding. With the development of internet technology and the ubiquitous presence of low-cost sur- veillance cameras nowadays, surveillance video has become a typical big data, offer- ing both opportunities and challenges for intelligent video surveillance. On one hand, the mass data involve more abundant information to mine. On the other hand, it suf- fers from various difficulties such as noise, label deficiency and computational com- plexity. This special session focuses on learning methods to achieve high performance video analysis and understanding under uncontrolled environments in large scale, which is also a very challenging problem. Moreover, it attracts much attention from both the academia and the industry. We hope this topic will aggregate top level works on the new advances in video analysis and understanding from big surveillance data.

Scope and Topics

We will solicit original contributions of researchers and practitioners from the aca- demia as well as industry, which address a wide range of theoretical and applied is- sues. The topics of interest include, but are not limited to:

  • Deep learning for wide area surveillance applications
  • Semi-supervised/unsupervised learning for surveillance
  • Learning for Camera topology construction and fusion
  • Video abstraction from mass surveillance data
  • Activity detection/recognition/profiling from mass data
  • Large surveillance video database annotation and searching
  • Object retrieval/identification for surveillance
  • Sensitive text, image/video discovery in surveillance data
  • Analysis and understanding the videos by unmanned aerial vehicle

IJCNN-11 Human-like Learning for Intelligent Systems

Organized by Cheng-Lin Liu and Zhaoxiang Zhang

Machine learning, with the aim of building intelligent systems by learning model or knowledge from data, has achieved great progress in the past 30 years. However, a huge gap of learning ability still exists between machine learning and human learning. For example, a five-year-old child can identify objects, understand speech and lan- guage via learning from small number of instances or daily communication, whereas machines can hardly match this ability even by learning from big data. In recent years, some researchers have attempted to develop machine learning methods simulating the human learning behavior. Such methods, called as “Human-like Learning”, have some features: learning from small supervised data, interactive, all-time incremental (life- long), exploiting contexts and the correlation between different data sources and tasks, etc. Some existing learning methods, such as incremental learning, active learn- ing, transfer learning, domain adaptation, learning with use, multi-task learning, zero- shot/one-shot learning, can be viewed as special/simplified forms of human-like learning. The future trend is to make learning methods more flexible and active, re- quiring less supervision, exploiting all kinds of data more adequately.

Scope and Topics

The topics of interest include, but are not limited to:

  • Brain-inspired neural networks
  • Human-like learning for deep models
  • Hybrid supervised and unsupervised learning
  • Learning from interaction
  • Learning with use
  • Zero/One-shot learning
  • Advanced transfer learning and adaptation
  • Advanced multi-task learning
  • Learning from heterogeneous data
  • Human-like learning for pattern recognition, computer vision, robotics and other applications

IJCNN-12 Deep Learning for Brain-Like Computing and Pattern Recognition

Organized by Guoqiang Zhong, Junyu Dong, Xinghui Dong, Hui Yu and Mohamed Cheriet

Deep learning is a topic of broad interest, both to researchers who develop new deep architectures and learning algorithms, as well as to practitioners who apply deep learning models to a wide range of applications, from image classification to video tracking, etc. Brain-like computing combines computational techniques with cognitive ideas, principles and models inspired by the brain for building information systems used in humans’ common life. Pattern recognition is a conventional area of artificial intelligence, which focuses on the recognition of patterns and regularities in data. Recently, there has been very rapid and impressive progress in these three areas, in terms of both theories and applications, but many challenges remain. This workshop aims at bringing together researchers in machine learning and related areas to discuss the utility of deep learning for brain-like computing and pattern recognition, the advances, the challenges we face, and to brainstorm about new solutions and directions.

Scope and Topics

A non-exhaustive list of relevant topics:

  • unsupervised, semisupervised, and supervised deep learning
  • active learning, transfer learning and multi-task learning
  • dimensionality reduction, metric learning and kernel learning
  • sparse modeling
  • ensemble learning
  • hierarchical architectures
  • optimization for deep models
  • intelligent data analysis and recommendation systems
  • implementation issues, parallelization, software platforms, hardware for deep learning and big data analysis
  • applications in video, image, texture, text processing, neuroscience, medical imaging or any other field

IJCNN-13 Advances in Computational Intelligence for Applied Time Series Forecasting (ACIATSF)

Organized by Cristian Rodriguez Rivero, Hector Daniel Patiño, Julian Antonio Pucheta, Gustavo Juarez and Leonardo Franco

Over the past few decades, application of simple statistical procedures with considerable heuristic or judgmental input was the beginning of forecasting, then in the 80’s, sophisticated time series models started to be used by some of the dynamic system operators, and these approaches, were to become pioneering works in this field.

Soft computing methods including support vectors regression (SVR), fuzzy inference system (FIS) and artificial neural networks (ANN) to time-series forecasting (TSF) has been growing rapidly in order to unify the field of forecasting and to bridge the gap between theory and practice, making forecasting useful and relevant for decision-making in many fields of the sciences.

The purpose of this session is to hold smaller, informal meetings where experts in a particular field of forecasting can discuss forecasting problems, research, and solutions in the field of automatic control. There is generally a nominal registration fee associated with attendance.

This session aims to debate in finding solutions for problems facing the field of forecasting. We wish to hear from people working in different research areas, practitioners, professionals and academicians involved in this problematic.

Scope and Topics

The session seeks to foster the presentation and discussion of innovative techniques, implementations and applications of different problems that are Forecasting involved, specially in real-world problems applied to control and automation.

  • Time Series Analysis
  • Time Series Forecasting
  • Evaluation of Forecasting Methods and Approaches
  • Forecasting Applications in Business, Energy and Price Demand, Hydrology and Rainfall
  • Impact of Uncertainty on Decision Making
  • Seasonal Adjustment
  • Multivariate Time Series Modelling and Forecasting
  • Marketing Forecasting
  • Economic and Econometric Forecasting

IJCNN-14 Advanced Supervised Learning Techniques and Its Applications

Organized by Min-Ling Zhang, Fuzhen Zhuang and Bing Han

Supervised learning techniques have been widely used in numerous real-world applications, ranging from information retrieval, multimedia content analysis, web mining, business intelligence to bioinformatics, research expedition, geological surveillance, public security, and so on. Traditional supervised learning makes several simplifying assumptions to facilitate the induction of learning systems, such as the strong supervision assumption that training examples carry sufficient and explicit supervision information, single domain assumption that training examples come from identical domain, uniform distribution assumption that training examples are class-balanced and have equal misclassification costs, etc.

Nonetheless, the above simplifying assumptions may not fully hold in practice due to various constraints imposed by physical environment, problem characteristics, and resource limitations. In recent years, researches on advanced supervised learning techniques have been rapidly growing to meet the increasing need in learning from data under non-trivial environments. The aim of this special session is to bring researchers and practitioners who work on various aspects of advanced supervised learning, to discuss on the state-of-the-art and open problems, to share their expertise and exchange the ideas, and to offer them an opportunity to identify new promising research directions.

Scope and Topics

This special session solicits papers whose topics fall into (but not limited to) the following categories:

  • Learning from labeled and unlabeled data, such as semi-supervised learning, PU learning, etc.
  • Learning from multi-instance data, such as multi-instance learning, multi-instance multi-label (MIML) learning, etc.
  • Learning from data with multiple class labels, such as multi-label learning, partial label learning, etc.
  • Learning from data from multiple domains, including transfer learning, multi-task learning
  • Learning from data with structured outputs
  • Applications of advanced machine learning techniques in biometric recognition
  • Applications of advanced machine learning in unstructured data analysis, including text, image, video, etc.
  • Applications of advance machine learning in cross-disciplinary fields, such as business intelligence, bioinformatics, space physics, astronomy, etc.
  • Applications of machine learning in big data, such as research expedition data, astronomy data, remote sensing data, etc.

IJCNN-15 Machine Learning with Incompletely Labeled Data

Organized by Yu-Feng Li, Sheng-Jun Huang and Min-Ling Zhang

Traditional supervised learning methods typically require the training data are fully labeled. Nowadays, the data size increases with an unprecedented speed. Fully labeled data becomes infeasible in many real situations, and consequently incomplete labeled data (or data with weak supervision) is ubiquitously existed. For years various approaches have been developed to learn with weak supervision, and learning from big data with weak supervision is showing its superiority to learning with fully labeled yet small data. However, there are still many open problems and in recent years many interesting challenges have been realized. For example, safe semi-supervised learning that prevents unlabeled data hurting the performance is desired; developing data-adaptive active learning strategies have not fully touched; effective partial label learning in the presence of class imbalance data; deriving high quality labels from noisy crowds; borrowing supervision from auxiliary sources, etc.

Scope and Topics

The main goal of this session is to provide a forum for researchers in this field to share the latest advantages in theories, algorithms, and applications on learning with incompletely labeled data. Authors are invited to submit their original work on learning with incompletely labeled data. The topics of interest include, but are not limited to:

  • Semi-supervised learning
  • Active learning
  • Partial label learning
  • Crowdsourcing
  • Multi-instance learning
  • Multi-instance multi-label learning
  • Learning with noisy labels
  • Weak label learning
  • Transfer learning
  • Zero-shot learning
  • Scalable or efficient learning algorithms

IJCNN-16 Applications of Machine Learning Techniques to Bioinformatics

Organized by Bin Liu

With the rapid development of advanced techniques in molecular biology, more and more sequence data has been generated, such as DNA, RNA, and protein sequences. A difficult and challenging task is how to discover useful knowledge based on the sequence data. To solve this problem, more and more machine learning techniques have been successfully applied, and valuable information has been uncovered by using these techniques, for example protein structure and function can be identified based on the protein sequences by using machine learning approaches, such as Artificial Neural network (ANN), Support Vector Machine (SVM), etc.

Scope and Topics

This special session focuses on exploring and applying advanced machine learning techniques to the field of bioinformatics. The topics include, but not limited to:

  • Protein structure and function identification
  • DNA binding protein identification based on supervised methods
  • Application of ranking algorithms to protein remote homology detection
  • Enhancer identification and classification
  • Promoter identification
  • MicroRNA identification

IJCNN-17 Complex-Valued Neural Networks

Organized by Yasuaki Kuroe, Tohru Nitta and Akira Hirose

The complex-valued neural networks (CVNNs) is a rapidly developing and growing area that has attracted continued interest for the last decade. The CVNN special session has become a traditional event of the IJCNN conference. Seven special sessions organized since 2006 (WCCI-IJCNN 2006, Vancouver, Porto, WCCI-IJCNN 2008, Hong Kong, IJCNN 2009, Atlanta, WCCI-IJCNN 2010, Barcelona, IJCNN-2011, San Jose, WCCI-IJCNN 2012, Brisbane, IJCNN-2013, Dallas, IJCNN-2014, Baijing) attracted numerous submissions and had large audiences. They featured many interesting presentations and very productive discussions.

There are several new directions in CVNN’s development: from formal generalization of the commonly used algorithms to the complex-valued case that are mathematically richer than regular neurons, to the use of original complex-valued activation functions that can increase significantly the neuron and network functionality. One of the new trends is the development of quaternion neurons and neural networks. There are also many interesting applications of CVNNs in pattern recognition and classification, nonlinear filtering, intelligent image processing, brain-computer interfaces, time series prediction, bioinformatics, robotics, etc.

One of the most important characteristics of the CVNNs is the proper treatment of phase and the information contained in phase, e.g., the treatment of wave-related rotation-related phenomena such as electromagnetism, light waves, quantum waves and oscillatory phenomena. Very interesting among other CVNNs are those networks that are based on neurons with the phase-dependent activation functions. This specific phenomenon makes it possible to increase the single neuron's functionality and to design more flexible and more efficient networks. It is also very interesting to study how the CVNNs can be used in modeling of the biological neural networks.

IJCNN 2016, which is an integrated part of IEEE WCCI 2016 will be a very attractive forum, where it will be possible to organize a systematic and comprehensive exchange of ideas in the area, to present the recent research results and to discuss the future trends. We hope that the proposed session will attract not only the potential speakers, but many new researches interested in joining the CVNNs community. We expect also that this session would be very beneficial for all computational intelligence researchers and other specialties that are in need of the sophisticated neural networks tools.

Scope and Topics

Papers that are, or might be, related to all aspects of the CVNNs are invited. We welcome contributions on theoretical advances as well as contributions of applied nature. We also welcome interdisciplinary contributions from other areas that are on the borders of the proposed scope. Topics include, but are not limited to:

  • Theoretical Aspects of CVNNs and Complex-Valued Activation Functions
  • Learning Algorithms for CVNNs
  • Complex-Valued Associative Memories
  • Pattern Recognition, Classification and Time Series Prediction using CVCNNs
  • CVNNs in Nonlinear Filtering
  • Dynamics of Complex-Valued Neurons
  • Learning Algorithms for CVCNNs
  • Chaos in Complex Domain
  • Feedforward CVCNNs
  • Spatiotemporal CVNNs Processing
  • Frequency Domain CVNNs Processing
  • Phase-Sensitive Signal Processing
  • Applications of CVNNs in Image Processing, Speech Processing and Bioinformatics
  • Quantum Computation and Quantum Neural Networks
  • CVNN in Brain-Computer Interfaces
  • CVNNs in Robotics
  • Quaternion and Clifford Networks

IJCNN-18 Neural Networks: Model and Application in Real World Problems

Organized by Zhi-Hui Zhan, Jing-Hui Zhong and Yong Wee Foo

Neural network (NN) has over 50 years of development and has been widely used as an efficient tool for many real-world applications. A typical NN contains the layers of neurons structure and the weights that connect the neurons. These structure and weight parameters can significantly impact the performance of NN. Although many learning algorithms have been proposed to optimize the NN parameters, they are still inadequate when dealing with complex real-world applications. The network modelling problem and optimization problem become more difficult and challenging nowadays not only because of the new and emerging NN paradigms such as deep learning networks which contain many neurons layers and weights, but also because of the difficulties in modelling the NN when dealing with challenging real-world applications in complex environments such as cloud and big data.

Therefore, when applying NN to real-world applications such as computer vision, natural language processing, speech recognition, classification, modeling and prediction, etc, in the cloud and big data era, how to model proper NN to present the problem and how to design approach for optimizing the NN structure and parameters are still open and significant research topics. Being powerful global optimization tool, evolutionary computation (EC) algorithms have fast developed in the past two decades and have also been widely applied to various optimization problems. The use of EC algorithms for optimizing the structure and parameters of NN reported in the literatures has shown promising performance although many open issues remain. This Special Session is to draw the attentions of researchers in both the communities of NN and EC to exchange their latest advances in theories and technologies of EC, NN, and the works on modeling the NN for real-world problems, designing the approaches for optimizing the NN, and extending the NN to real-world applications.

Scope and Topics

Evolutionary computation (EC) algorithms contains evolutionary algorithms (EAs) such as genetic algorithm (GA), evolution strategy (ES), evolutionary programming (EP), estimation of distribution algorithm (EDA), differential evolution (DE), etc, and swarm intelligence (SI) algorithms such as ant colony optimization (ACO), particle swarm optimization (PSO), artificial bee colony (ABC), brain storm optimization (BSO), etc. These EC algorithms are efficient in optimizing both continuous optimization problem and/or discrete combinational optimization problems. The neural network (NN) optimization can be modelled as an optimization problem suitable for EC algorithms. This special session mainly focuses on the researches of NN Model, Approach, and Application. Works on modelling of new NN paradigms, especially based on the real-world problems, are welcome. Works on designing new optimization approaches, especially on using EC algorithms to optimize the NN structure and/or weight, are welcome. Moreover, the real-world applications of NN and evolutionary NN (ENN) are welcome. Authors are invited to submit their original and unpublished work with the topics including, but not limited to:

  • Modelling of neural networks paradigm;
  • Modelling of neural networks paradigm based on real-world application problems;
  • Approaches for optimizing neural networks;
  • Approaches for optimizing neural networks based on evolutionary computation algorithms;
  • Theoretical analysis of neural networks;
  • Theoretical analysis of evolutionary neural networks;
  • Real-world applications of neural networks;
  • Real-world applications of evolutionary neural networks;
  • Other related to EC or NN.

IJCNN-19 Ordinal Regression and Ranking

Organized by Pedro A. Gutiérrez, María Pérez­Ortiz and Peter Tiňo

Ordinal regression (or ordinal classification) is a relatively new learning problem, where the objective is to learn a rule to predict labels in an ordinal scale­ discrete labels endowed with a natural order. Consider, for example, the case of a teacher who rates student's performance using A, B, C, D and E, and A>B>C>D>E. Such order information could be helpful for constructing more robust and fair classifiers and evaluation metrics. On the other hand, ranking generally refers to those problems where the algorithm is given a set of ordered labels, and the objective is to learn a rule to rank patterns by using this discrete set of labels. Many real problems exhibit this structure, e.g. multi­criteria decision making, medicine, risk analysis, university ranking, information retrieval and filtering.

Specific solutions have been recently proposed in the machine learning and pattern recognition literature for both ordinal regression and ranking problems, resulting in a very active research field. This special session aims to cover a wide range of approaches and recent advances in ordinal regression and ranking. We hope that this session can provide a common forum for researchers and practitioners to exchange their ideas and report their latest findings in the area.

Scope and Topics

In particular we encourage submissions addressing the following issues:

  • Extensions of standard classification methods to ordinal regression (Support Vector Machines, Gaussian Processes, Discriminant Analysis, etc.).
  • Threshold models for ordinal regression.
  • Decomposition methods for ordinal regression and ranking.
  • Imbalanced ordinal regression problems.
  • Clustering and pre­processing methods for ordinal and monotonic data (data cleaning techniques, feature selection, etc).
  • Evaluation measures for ordinal regression and ranking.
  • Learning to rank: ranking (pointwise, pairwise and listwise algorithms), sorting and multipartite ranking.
  • Monotonic classification methods.
  • Preference learning.
  • Applications in medicine, information retrieval, recommendation systems, risk analysis and any other real­world problems.

IJCNN-20 Computational Intelligence Algorithms for Digital Audio Applications

Organized by Stefano Squartini, Aurelio Uncini, Björn Schuller and Francesco Piazza

Computational Intelligence (CI) techniques are largely used to face complex modelling, prediction, and recognition tasks in different research fields. One of these is represented by Digital Audio, which finds application in entertainment, security, forensics and health. Anyone can experience a large variety of servies and products including Digital Audio technologies, undoubtedly characterized by a progressively increase of complexity, interactivity and intelligence.

The typical methodology adopted in these engineering solutions consists in extracing and manipulating useful information from the audio stream of pilot the execution of automatized services. Several technical areas in Digital Audio, involving different kinds of audible signals, share such an approach. In the "music" case study, music information retrieval is the major topic to addresss, with many diverse sub-topics therein; for "speech", we can immediately refer to speech/speaker recogntion, but also the many diverse topics intimately related to the computational analysis of speech signals ( affective computing and spoken language processing, just to name a few); in the case of "sound", acousic fingerprint/signature, acoustic monitoring and sound detection/identification have lately seen an ever increasing interest in the larger field. Moreover, also cross-domain approaches to exploit the information contained in diverse kinds of environmental audio signals have recently been investigated. In many application contexts, this appears in conjunction with data coming from other media, such as textual and visual data, for which specific fusion techniques are required.

In dealing with these problems, the adoption fo data-driven learning systems is often a "msut", and the recent success encountered by deep neural architectures (in Speech Recongnition, for instance) lends further evidence of this. inherent challenges are, however coming with technological issues, due to the presence of non-stationary operating conditions and hard real-time constraints, often made harder by the big amount of data to process. In some other application contexts, the challenge is facing a scarce amount of data to be used for training, and suitable architectures and algorithms need to be designated on purpose. Last but not least, a key issue in Intelligent Audio Applications is given by the capability to learn represntative features at different abstraction layers without the support of supervised actions. Again, the deep learning paradigm recently allowed reaching relevant achievements in this sense, with many open issues to investigate.

It is indeed of great interst for the scientific community to understand how and to what extent novel Computational Intelligence-based techiques (with special attention to Neural Network ones) can be efficiently employed in Digital Audio, in the light of all the aforementioned aspects. The aim of the session is therefore to focus on the most recent advancements in the Computational Intelligence field and on their applicability to Digital Audio problems. Driven by the sucess encountered at IJCNN2014 in Bejing (China) and IJCNN2015 in Killarney (Ireland), the proposers of this session are highly motivated to revive and exceed the experience and to build, in the long-term, a solid reference within the Computaional Intelligence community for the Digital Audio field.

Scope and Topics

Topics include, but are not limited to:

  • Computational Audio Analysis
  • Deep Learning algorithms in Digital Audio
  • Neural Architectures for Audio Processing
  • Transfer, Weakly Supervised, and Reinforcement Learning for Audio
  • Music Information Retrieval
  • Speech and Speaker Analysis and Classification
  • Sound Detecion and Identification
  • Acoustic Source Separation
  • Acooustic Novelty Detection
  • Computational methods for Wireless Acoustic Sensor Networks
  • Brain inspired Auditory Scene Analysis
  • Cross-domain Audio Analysis
  • Speech and Audio Forensics
  • Audio-based Security Systems
  • Intelligent Audio Interfaces

IJCNN-21 Deep Learning, Medical Imaging, and Translational Medicine

Organized by Qian Wang, Jun Shi, Shihui Ying, Manhua Liu and Yonghong Shi

Deep learning has demonstrated its capability for many vision problems, such as face detection and recognition, image classification, etc. It is expected that this technique can benefit the area of medical image analysis, as well as imaging-based translational medicine. Though a few pioneering works can be found in the literature, there are still a lot of unresolved issues when applying deep learning for medical images.

The goal of special session is to present works that focus on the design and use of deep learning in medical image analysis as well as imaging-based translational medical studies. This special session is going to set the trends and identify the challenges of the use of deep learning methods in the field of medical image. Meanwhile, it is expected to increase the connection between software developers, specialist researchers and applied end-users from diverse fields.

Scope and Topics

Topics include, but are not limited to:

  • Image descriptor and feature extraction
  • Image super-resolution
  • Image reconstruction
  • Image registration
  • Image segmentation and labeling
  • Computer-assisted lesion detection
  • Computer-assisted diagnosis
  • Deep learning model selection
  • Meta-heuristic techniques for fine-tuning parameter in deep learning-based architectures
  • Other related translational medical applications.

IJCNN-22 Sequential Learning with Neural Networks

Organized by Huanhuan Chen, Giacomo Boracchi and Jian Cheng

Over the past few decades, research and applications of sequential data has attracted growing attention from both scientific and industrial communities. A number of processing techniques have been proposed for sequential data understanding and processing, e.g. dynamic time warping (DTW), fisher kernel and recurrent neural networks. The main aim of this special session is not only to explore the new techniques on this area, providing original research with the aim for deeper understanding into the mechanism of algorithms, but also to encourage exchange of great ideas on sequential learning in different scenarios. We wish to communicate with people working in different research areas, practitioners, professionals and academicians in this area.

Scope and Topics

Topics include, but are not limited to:

  • Recurrent neural networks
  • Sequential data classification, regression and learning
  • Architectures, techniques and algorithms for learning in non-stationary/dynamic environments
  • Domain adaptation, dataset shift, covariance shift
  • Incremental learning, lifelong learning, cumulative learning
  • Change-detection tests and anomaly-detection algorithms
  • Mining from streams of data
  • Sequential kernel
  • Text Classification
  • Streaming learning and mining
  • Hidden Markov models
  • Echo state networks, and reservoir models
  • Applications that call for incremental learning or learning in non-stationary/dynamic environments, such as:
    • Adaptive classifiers for concept drift and recurring concepts
    • Intelligent systems operating in non-stationary/dynamic environments
    • Intelligent embedded and cyber-physical systems
  • Applications that call for change and anomaly detection, such as:
    • Fault detection
    • Fraud detection
    • Network intrusion and security
    • Intelligent sensor networks
  • Cognitive-inspired approaches to adaptation and learning
  • Finite state machines and Dynamic models
  • Issues relevant to above mentioned or related fields

IJCNN-23 Research on Large Scale Clustering Algorithms

Organized by Yiming Zhang, Shaohe Lv, Xin Niu and Xinwang Liu

This special session aims to promote new advances and research directions to address the clustering problem in large scale practical applications.
Unprecedented technological advances lead to increasingly large scale data sets in all areas of science, engineering and businesses. These include genomics and proteomics, biomedical imaging, signal processing, astrophysics, finance, web and market basket analysis, among many others. The number of such data is often of the order of millions or billions. Classical clustering algorithms become inadequate, questionable, or inefficient at best, and this calls for new clustering algorithms.

Scope and Topics

Topics of interest include theoretical foundations, algorithms and implementation, as well as applications and empirical studies, for example:

  • Systematic studies of how the number of samples affects clustering algorithms
  • New clustering algorithms
  • Research on parallel clustering algorithms
  • Multiple clustering algorithm ensemble
  • Methods of random projections, compressed sensing, and random matrix theory applied to large scale clustering
  • Theoretical underpinning of large scale clustering algorithms
  • Data presentation and visualisation methods for large scale datasets
  • Applications to real problems in science, engineering or businesses where the data is large

IJCNN-24 Deep Learning for Big Multimedia Understanding

Organized by Jinhui Tang and Zechao Li

Conventional multimedia understanding is usually built on top of handcrafted features, which are often much restrictive in capturing complex multimedia content. Recent progress on deep learning opens an exciting new era, placing multimedia understanding on a more rigorous foundation with automatically learned representations to model the multimodal data and the cross-media interactions. Existing studies have revealed promising results that have greatly advanced the state-of-the-art performance in a series of multimedia research areas, from the multimedia content analysis, to modeling the interactions between multimodal data, to multimedia content recommendation systems, to name a few here.

Scope and Topics

This special session aims to provide a forum for the presentation of recent advancements in deep learning research that directly concerns the multimedia community. For multimedia research, it is especially important to develop deep learning methods to capture the dependencies between different genres of data, building joint deep representation for diverse modalities. The list of topics includes and is not restricted to the following:

  • Novel deep network architectures for multimodal data representation
  • Deep learning for new multimedia applications
  • Efficient training and inference methods for multimedia deep networks
  • Emerging applications of deep learning in multimedia search, retrieval and management
  • Deep learning for multimedia content analysis and recommendation
  • Deep learning for cross-media analysis, knowledge transfer and information sharing
  • Distributed computing, GPUs and new hardware for deep learning in multimedia research
  • Other deep learning topics for multimedia computing, involving at least two modalities

IJCNN-25 Distributed Learning Algorithms for Neural Networks

Organized by Massimo Panella and Simone Scardapane

In the era of big data and pervasive computing, it is common that datasets are distributed over multiple and geographically distinct sources of information (e.g. distributed databases). In this respect, a major challenge is designing adaptive training algorithms in a distributed fashion, with only partial or no reliance on a centralized authority. Indeed, distributed learning is an important step to handle inference within several research areas, including sensor networks, parallel and commodity computing, distributed optimization, and many others. Additionally, it generalizes previous research on training neural and fuzzy neural models over clusters of processors and, as such, it is crucial in designing training algorithms for efficiently processing large amount of data over networks.

Based on the idea that all the aforementioned research fields share many fundamental questions and mechanisms, this special session is intended to bring forth advances on distributed training for neural networks. We are interested in papers proposing novel algorithms and protocols for distributed training under multiple constraints, analyses of their theoretical aspects, and applications for multiple source data clustering, regression and classification.

Scope and Topics

The topics of interest to be covered by this Special Session include, but are not limited to:

  • Distributed algorithms for training neural networks
  • Theoretical aspects of distributed learning (e.g. fundamental communication constraints)
  • Learning on commodity computing architectures and parallel execution frameworks (e.g. MapReduce, Storm)
  • Energy efficient distributed learning
  • Distributed semi-supervised and active learning
  • Novel results on distributed optimization for machine learning
  • Cooperative and competitive multi-agent learning
  • Learning in realistic wireless sensor networks
  • Distributed systems with privacy concerns (e.g. healthcare systems)

IJCNN-26 Concept Drift, Domain Adaptaion and Learning in Dynamic Environments

Organized by Giacomo Boracchi and Robi Polikar, Manuel Roveri and Gregory Ditzler

Learning abilities of computational models have been well-researched with promising progress, but a vast majority of these efforts still rely on two fundamental assumptions:

  • there is a sufficient and representative data set;
  • data are drawn from a fixed – albeit unknown – distribution.
Alas, these assumptions often do not hold in many applications of practical importance. Recently, concept-drift and domain-adaptation algorithms have attempted to remove this assumption, operating on datastreams whose underlying distribution changes over time. These approaches, however, are still primarily of heuristic nature, with many parameters requiring fine-tuning, and have not been evaluated on large scale real-world applications, yet.

Considering that the ultimate goal of computational intelligence is to attain brain-like intelligence and that the plasticity of brain-like intelligence can, and routinely does, learn incrementally and in dynamic (nonstationary) environments, the need for a framework for learning in non- stationary environment is very real. Moreover, considering the growing number of real-world applications that can benefit from these algorithms, such as learning from financial/environmental data or predicting user interests (to name a few examples), it is clear that there is much work to be done for solving the problem of learning in non-stationary environment.

The proposed session has three main goals:
  • Disseminate recent developments concerning learning in dynamic environments and discuss results concerning both the theoretical perspective of machine learning and the application- oriented view of computational intelligence.
  • Provide a forum for researchers in this area to exchange new ideas with each other, as well as with the rest of the neural network & computational intelligence community.
  • Illustrate relevant applications involving learning in dynamic environments and discuss the next challenges in this field.

Scope and Topics

Researchers working in any of the related areas of learning in dynamic/nonstationary environments, concept drift or domain adaptation are encouraged to submit their contributions to this special session.
The scope of the proposed session includes, but is not limited to:

  • Methodologies, algorithms and techniques for learning in dynamic/non-stationary environments
  • Incremental learning, lifelong learning, cumulative learning
  • Domain adaptation and dataset-shift, covariate-shift adaptation
  • Semi-supervised learning methods for handling concept-drift
  • Ensemble methods for learning under concept drift
  • Learning under concept drift and class unbalance
  • Change-detection tests and anomaly-detection algorithms
  • Algorithms for information mining in nonstationary datastreams
  • Applications that call for learning in dynamic/non-stationary environments, and for incremental learning, such as:
    • Adaptive classifiers for concept drift and recurring concepts
    • Intelligent systems operating in dynamic/non-stationary environments
    • Intelligent embedded and cyber-physical systems
  • Applications that call for change and anomaly detection, such as:
    • fault detection
    • fraud detection
    • network-intrusion detection and security
    • intelligent sensor networks
  • Cognitive-inspired approaches to adaptation and learning
  • Development of test-sets benchmarks for evaluating algorithms learning in non-stationary/dynamic environments
  • Issues relevant to above mentioned or related fields

IJCNN-27 Extreme Learning Machines (ELM)

Organized by Guang‐Bin Huang, Jonathan Wu and Donald C. Wunsch II

Over the past few decades, conventional computational intelligence techniques faced bottlenecks in learning (e.g., intensive human intervention and time consuming). With the ever increasing demand of computational power particularly in areas of big data computing, brain science, cognition and reasoning, emergent computational intelligence techniques such as extreme learning machines (ELM) offer significant benefits including fast learning speed, ease of implementation and minimal human intervention.

Extreme Learning Machines (ELM) aim to break the barriers between the conventional artificial learning techniques and biological learning mechanism. ELM represents a suite of machine learning techniques for (single and multi‐) hidden layer feedforward neural networks in which hidden neurons need not be tuned. From ELM theories point of view, the entire multilayers of networks are structured and ordered, but they may be seemingly ‘‘messy’’ and ‘‘unstructured’’ in a particular layer or neuron slice. ‘‘Hard wiring’’ can be randomly built locally with full connection or partial connections. Coexistence of globally structured architectures and locally random hidden neurons happen to have fundamental learning capabilities of compression, feature learning, clustering, regression and classification. ELM theories also give theoretical support to local receptive fields in visual systems.

ELM learning theories show that hidden neurons (including biological neurons whose math modelling may be unknown) (with almost any nonlinear piecewise activation functions) can be randomly generated independent of training data and application environments, which has recently been confirmed with concrete biological evidences. ELM theories and algorithms argue that “random hidden neurons” capture the essence of some brain learning mechanism as well as the intuitive sense that the efficiency of brain learning need not rely on computing power of neurons. This may somehow hint at possible reasons why the brain is more intelligent and effective than computers. ELM offers significant advantages such as fast learning speed, ease of implementation, and minimal human intervention. ELM has good potential as a viable alternative technique for large‐scale computing and artificial intelligence.

The need for efficient and fast computational techniques poses many research challenges. This special session seeks to promote novel research investigations in ELM and related areas.

Scope and Topics

All the original papers related to ELM technique are welcome. Topics of interest include but are not limited to:
Theories

  • Universal approximation, classification and convergence
  • Robustness and stability analysis
  • Biological learning mechanism
Algorithms
  • Real‐time learning, reasoning and cognition
  • Sequential/incremental learning and kernel learning
  • Clustering and feature extraction/selection
  • Random projection, dimensionality reduction, and matrix factorization
  • Closed form and non‐closed form solutions
  • Multi hidden layers solutions, hierarchical ELM, and random networks
  • No‐Prop, Random Kitchen Sink, FastFood, QuickNet, RVFL
  • Parallel and distributed computing / cloud computing
Applications
  • Time series prediction
  • Pattern recognition
  • Web applications
  • Biometrics and bioinformatics
  • Power systems and control engineering
  • Security and compression
  • Human computer interface and brain computer interface
  • Cognitive science/computation
  • Sentic computing / natural language processing
  • Data analytics, super / ultra large‐scale data processing

IJCNN-28 Biologically-inspired Neural Networks for Robotics

Organized by Chaomin Luo

Biologically-inspired intelligence technique, an important embranchment of series on computational intelligence, plays a crucial role for robotics. The autonomous robot and vehicle industry has had an immense impact on our economy and society, and this trend will continue with biologically inspired neural networks techniques. Biologically-inspired intelligence, such as biologically-inspired neural networks (BNN), is about learning from nature, which can be applied to the real world robot and vehicle systems. Recently, the research and development of bio-inspired systems for robotic applications is increasingly expanding worldwide. Biologically-inspired algorithms contain emerging sub-topics such as bio-inspired neural network algorithms, brain-inspired neural networks, swam intelligence with BNN, ant colony optimization algorithms (ACO) with BNN, bee colony optimization algorithms (BCO), particle swarm optimization with BNN, immune systems with BNN, and biologically-inspired evolutionary optimization and algorithms, etc. Additionally, it is decomposed of computational aspects of bio-inspired systems such as machine vision, pattern recognition for robot and vehicle systems, motion control, motion planning, movement control, sensor-motor coordination, and learning in biological systems for robot and vehicle systems.

This special session seeks to highlight and present the growing interests in emerging research, development and applications in the dynamic and exciting areas of biologically-inspired algorithms for robot and vehicle systems (autonomous robots, unmanned underwater vehicles, and unmanned aerial vehicles).

Scope and Topics

Original research papers are solicited in related areas of biologically-inspired algorithms for robotics. Submissions to the Special Session should be focused on theoretical results or innovative applications of computational intelligence of biologically-inspired algorithms (such as BNN) for robot and vehicle systems.

Specific topics for the special session include but are not limited to:

  • Biologically-inspired neural networks for robotics
  • Artificial neural networks and learning systems for robotics such as motion planning, navigation, mapping, localization, image processing, etc
  • Bio-inspired system on computer vision and image progressing for robotics
  • Human-like learning for robotics
  • Theory, design, and applications of neural networks and related learning systems for robotics and vehicles
  • Evolutionary optimization, machine vision, pattern recognition for robot and vehicle systems
  • Brain-inspired neural networks for robotics
  • Swarm intelligence for robotics
  • Evolutionary neuro-computing for robot and vehicle systems
  • Bio-inspired system on machine learning, intelligent systems design for robotics
  • Cellular automata for robotics
  • Immune systems with BNN for robotics
  • Ant colony optimization algorithms (ACO) with BNN for robotics
  • Bee colony optimization algorithms (BCO) with BNN for robotics.

IJCNN-29 Reinforcement Learning and Approximate Dynamic Programming for Optimization in Dynamic Environment

Organized by Daoyi Dong, Dongbin Zhao and Qinmin Yang

Reinforcement learning and approximate dynamic programming can be used to address learning and optimization problems in many areas of engineering and science, including artificial intelligence, control engineering, operation research, psychology, and economy. They have provided critical tools to solve some engineering and science problems in modern complex systems. However, there still exist some challenges in the applications of reinforcement learning and approximate dynamic programming to academic and industrial problems such as the curse of dimensionality and optimization in dynamic environment. At the same time, the development of new technologies such as quantum technology and deep learning provides a remarkable opportunity to revisit these challenges in reinforcement learning and approximate dynamic programming. This special session will focus on relevant topics of reinforcement learning and approximate dynamic programming, and provide a forum for idea exchange in the emerging research area.

Scope and Topics

The aim of this special session will be to provide an account of the state-of-the-art in this fast moving and cross-disciplinary field of reinforcement learning and approximate dynamic programming. It is expected to bring together the researchers in relevant areas to discuss latest progress, propose new research problems for future research. All the original papers related to reinforcement learning (RL) and approximate dynamic programming (ADP) are welcome. Topics of interest include but are not limited to:

  • ADP for discrete-time systems
  • ADP for continuous-time systems
  • Policy iteration algorithm
  • Hierarchical reinforcement learning
  • Multi-agent reinforcement learning
  • Deep reinforcement learning
  • Quantum reinforcement learning
  • Applications of ADP and RL to optimization in dynamic environment

IJCNN-30 Biologically Inspired Computational Vision

Organized by Khan M. Iftekharuddin

Constructive understanding of computational principles of visual information processing, perception and cognition is one of the most fundamental challenges of contemporary science. Deeper insight into biological vision helps to advance intelligent systems research to achieve robust performance similar to biological systems. Biological inspiration indicates that sensory processing, perception, and action are intimately linked at various levels in animal vision. Implementing such integrated principles in artificial systems may help us achieve better, faster and more efficient intelligent systems. This session provides an integrated platform to present original ideas, theory, design, and applications of computational vision.

Scope and Topics

Topics of interest include, but are not limited to the following:

  • Theoretical approaches and modeling in computational vision
  • Neuronal mechanisms of visual processing
  • Low level vision and its relationship to biological machinery
  • Artificial learning systems for image and information processing and evidential reasoning for recognition
  • Intelligent search in communications networks
  • Modeling issues in ATM networks, agent-oriented computing architectures
  • Perception of shape, shadows, poses, color and illumination in object recognition
  • Tracking for inferring shapes and 3D motions
  • Active visual perception, attention and robot vision
  • Functional Magnetic Resonance Imaging (fMRI) studies of visual segmentation and perception
  • Application of computational vision in areas of
    • Automated target identification and acquisition systems in defense and industry
    • Biomedical imaging
    • 3D photography
    • Face recognition
    • Learning to segment camouflaged objects
    • Motor actions and robotics
    • Image databases and indexing
  • Hardware implementation of computational vision
  • Any other topics related to biological approaches in computer vision

IJCNN-31 Privacy-preserving in Data Mining and Applications

Organized by Tianqing Zhu,, Gang Li and Ping Xiong

Over the past two decades, digital information collected by corporations, organizations and governments has resulted in huge number of datasets, and the speed of such data collection has increased dramatically over the last a few years. A data collector, also known as a curator, is in charge of releasing and publishing data for data mining and some applications. However, most of the collected datasets are personally related and contain private or sensitive information. Even though curators can apply several simple anonymization techniques, there is still a high probability that the sensitive information of individuals has a high probability to be disclosed. Privacy-preserving has therefore become an urgent issue that needs to be addressed.

Research communities have proposed various methods to preserve privacy and have designed a number of metrics to evaluate the privacy level of these methods. The interest in this area is very high and the notion is spanning in a range of research areas, ranging from the privacy community, to the data science communities including machine learning, data mining, statistics and learning theory. Much work has been conducted in a number of application domains, including social network, online education, recommender systems and tourism. A significant number of new technologies and applications have appeared in the privacy-preserving research area. We believe that it is a good time to cover these topics in the special session, which will include recent advances and applications in the privacy preserving research in data mining and diverse applications.

Scope and Topics

All submissions will be rigidly peer reviewed to guarantee the quality. This special session will focus on original articles in relevant topics, which include but are not limited to:

  • Privacy-preserving in applications and services
  • Privacy-preserving in Big Data
  • Privacy-preserving data analysis
  • Privacy-preserving in database systems
  • Privacy-preserving in cloud/mobile/wireless communications
  • Privacy-preserving in E-learning and e-government
  • Miscellaneous privacy issues

IJCNN-32 Advanced Machine Learning Methods and Applications from Complicated Data Environment

Organized by Jia Wu, Shirui Pan, Peng Zhang, Xingquan Zhu, Chengqi Zhang and Philip S. Yu

Traditional machine learning methods have been commonly used for many applications, such as text classification, image recognition, video tracking, etc. For learning purposes, these data are often required to be represented as vectors. However, many other objects in real-world applications, such as chemical compounds in bio pharmacy, brain regions in brain network and users in social network, contain rich feature vectors and structure information. Such a simple feature-vector representation inherently loses the structure information of the objects. In reality, objects may have complicated characteristics, depending on how the objects are assessed and characterized. Meanwhile the data may come from heterogeneous domains, such as traditional tabular-based data, sequential patterns, social networks, time series information, and semi-structured data. For the purpose of preserving the information accommodate complicated characteristics of the objects to adapt to the advanced applications, novel machine learning methods are desired to learn and discover the meaningful knowledge.

Scope and Topics

This special session expects to solicit contributions on the advanced machine learning methods and applications from complicated data environment. The topics of interest include, but are not limited to:

  • Semi-structured Learning
  • Graph-based Learning
  • Graph Classification/Clustering/ Streaming
  • Multi-Graph Learning
  • Deep Graph Learning
  • Online Graph Learning
  • Time Series Learning
  • Complex Social Networks
  • Multi-view/instance/ label Learning
  • Heterogeneous Transfer Learning
  • Web/Text/Image Mining
  • Multimedia Learning

IJCNN-33 Computational Intelligence and Machine Learning for Power Grid Security

Organized by Zhen Ni, Haibo He, Yan Sun, and Dongbin Zhao

Through the recent years, new computational intelligence and machine learning methodologies and frameworks have been developed as useful techniques to address power grid security issues. The blackouts, cascading failures, cyber/physical attacks and other stability issues have attracted researchers from cross-discipline to look into this interdisciplinary topic. Modern and complex power grid operation and safety issues also need the special attention and involvement of people with expertise in aforementioned fields. This special session will provide a unique platform for researchers from different societies, including computational intelligence, machine learning, power and energy, cyber security, communications, neuroscience and among others, to share their research experience towards a secure and smart modern power grid. The special session will also enhance the discussion among different communities to explore more challenge cross-discipline topics along this direction.

Scope and Topics

This special session will provide a forum to discuss recent development in power grid security based on all kinds of computational intelligence and machine learning techniques. We are particularly interested in the following topics:

  • Computational intelligence for cascading failures in power system
  • Computational intelligence for multi-contingency analysis in power system
  • Computational intelligence for power grid security alarm forecasting, planning and assessment
  • Computational Intelligence for cyber and/or physical attack in power system
  • Computational Intelligence for power grid risk assessment
  • Machine learning for power system data recognition, detection, classification/clustering and analysis
  • Machine learning for power system intrusion detection
  • Machine learning for power grid transmission and communication security (i.e., encryption, authentication and access control)
  • Machine learning for power grid risk modeling (e.g., vulnerability assessment, impact analysis and trust management)
  • Power grid (e.g, generation, transmission and load) vulnerability Assessment and physical security
  • AC/DC power flow stability analysis based on computational intelligence and machine learning

IJCNN-34 Smart Educational Techniques in Big Data Age

Organized by Guandong Xu, Gang Li and Wu He

Data mining provides educational institutions the capability to explore, visualize and analyze large amounts of data in order to reveal valuable patterns in students' learning behaviors without having to resort to traditional survey methods. Turning raw data into useful information and knowledge also enables educational institutions to improve teaching and learning practices, and to facilitate the decision-making process in educational settings. Thus, it is becoming important for researchers to exploit the abundant data generated by various educational systems for enhancing teaching, learning and decision making.

In addition, the development and training of teachers in regional area can be also improved by adopting smart techniques in the big data age. How to get these tasks done smartly and effectively is another important issue that could be potentially addressed in Big data age.

Scope and Topics

All submissions will be rigidly peer reviewed to guarantee the quality. This special session will focus on original articles in relevant topics, which include but are not limited to:

  • Data mining in education
  • Learning analytics
  • "Big Data" applications and opportunities in learning and education
  • Integrating data mining and pedagogical theory
  • Data mining with emerging pedagogical environments such as educational games and MOOCs
  • Recommender systems for learning
  • Case studies in Educational Data
  • Data Driven Performance Evaluation

IJCNN-35 Spiking Neural Networks

Organized by Nathan Scott and Nikola Kasabov

Spiking Neural Networks are a rapidly emerging means of neural information processing, drawing inspiration from biological processes. There is presently considerable interest in this topic, especially with the recent announcement of large scale projects such as the “BRAIN Initiative” (US) and the “Human Brain Project” (EU). Due to their inspiration from human brain processes, SNN have the potential to advance technologies and techniques in fields as diverse as medicine, finance, computing, and indeed any field that involves complex spatio-temporal data. SNN can operate on noisy data, in changing enviroment,s at low power and with high effectivness. We believe that this area is quickly establishing itself as an effective alternative to traditional machine learning technologies, and that interest in this area of research is growing rapidly. In this special session we intend to provide a platform for the discussion of contemporary areas of SNN, including theory, applications, and emerging technologies such as neuromorphic hardware.

Scope and Topics

Topics of interest include, but are not limited to the following:

  • Novel architectures
  • SNN applications and case studies
  • Learning algorithms for SNN, including Deep Learning
  • Big data and stream data processing in SNN
  • Theory or practice in biologically realistic neural simulation or biomimetic models
  • Neuromorphic hardware systems and applications
  • Robotic applications of SNN
  • Theory of SNN
  • Optimisation of SNN
  • Evolving SNN
  • Any other topics relating to Spiking Neural Networks, their theory, or application.

IJCNN-36 Online Real-Time Learning Strategies for Large Data Streams

Organized by Mahardhika Pratama, Meng Joo Er, Edwin Lughofer, Wenny Rahayu, Chee-Peng Lim and Ning Wang

Learning from large data streams is a research area of growing interest because large volumes of data are continuously generated from sensors, the Internet, etc., at an increasing high rate. The major difficulty of online learner in learning from large data streams results from uncertainties arising from three causes, namely real-time situations, data distribution and data representation:

  1. Uncertainties in a real-time situation: Since data streams may be very large and even possibly unbounded, learning from data streams usually requires a real-time algorithm through a single-pass or incremental learning scenario. Uncertainties in the real-time situations clearly occur in an incremental manner so as to achieve a trade-off between emphasizing new observations and existing knowledge. This is the well-known plasticity-stability dilemma.
  2. Uncertainties in data distribution: Uncertainties in the data distribution reflect a situation in which the input and/or output concepts do not follow static and predictable data distributions, but may change over time. This issue is also wellknown as concept drift. In the classification problem, the uncertainties in data distribution can also be perceived in the existence of class overlapping, which leads to a confusion of classifiers in classifying data points.
  3. Uncertainties in data representation: Uncertainties in data representation are attributed by inexact, inaccurate and uncertain characteristics of data themselves. This phenomenon occurs as a result of disagreements in expert knowledge, noisy measurements, or sparse data.
This special session aims to bring together research works of online real-time learning strategies for large data streams. Special attention will be devoted to handle the issue of uncertainty come across by online learner in learning large data streams.

Scope and Topics

The main topics of this special session include, but are not limited to, the following:

  • Online real-time unsupervised learning and clustering for large data streams
  • Online real-time supervised classification and regression for large data streams
  • Online real-time time-series modelling for large data streams
  • Online real-time intelligent controller for large data streams
  • Appropriate handling of data uncertainty in learning from large data streams
  • Tools and techniques for data stream mining in uncertain environments
  • Computational intelligence methods for big data analytics and huge data bases
  • Techniques to address drifts and shifts in data streams
  • On-line dynamic dimension reduction in high-dimensional streams
  • Feature selection and extraction techniques for large data streams
  • Sample selection and active learning for large data streams
  • Robustness and safety aspect in learning from large data streams
  • Practical applications of computational intelligence techniques for data stream mining
  • Real world cases of uncertainties in large data stream

IJCNN-37 Approximate/Adaptive Dynamic Programming for Control of Cyber-physical Systems

Organized by Ali Heydari and Kyriakos Vamvoudakis

Cyber-physical systems (CPS), where dynamical systems are subject to computation and/or communication considerations are emerging in different fields, from aerospace and manufacturing to healthcare and economics. Control of CPS is different from traditional control due to the required incorporation of the cyber part of the systems. The part gives rise to concerns about stability and performance of the system due to issues including communication delays, information losses, limited communication bandwidths, quantization errors, limited computational resources, and vulnerability to cyberattacks. These emerging challenges call for new methods and tools for effective control of CPS. Approximate/adaptive dynamic programming (ADP) as a powerful scheme for approximating solutions to optimal control problems has shown many potentials in solving this class of problems, in recent years.

Scope and Topics

This special session is aimed at providing a forum for presenting latest developments in the field (control of CPS) using the tool (ADP). Topics of interest include, but are not limited to:

  • Control under random delays and packet losses
  • Control under quantization errors
  • Event triggered control
  • Co-design of triggering and control policies
  • Decentralized control of CPS
  • Cyber-security in CPS
  • Fault detection in CPS
  • Robust control of CPS
  • Game Theory
  • Autonomy
  • Applications of CPS in transportation, manufacturing, healthcare, etc.

IJCNN-38 Energy Efficient Deep Neural Networks

Organized by Andrew Cassidy, Alexander Andreopoulos, Michael DeBole and Arnon Amir

Recent algorithmic advances in deep neural networks have made enormous strides in accuracy on a wide range of applications and domains (images, video, audio, speech, natural language, etc). At the same time, emerging work in low-power computing platforms, inspired by the structure of the brain, have demonstrated orders of magnitude improvement in computational efficiency and have scaled up to that point where they can address real-world machine learning challenges. Combining these two disciplines has the potential to enable truly revolutionary solutions and applications, ranging from small-scale embedded mobile systems up to very large-scale scientific and data center installations.

The objective of this special session is to report on work at this intersection, applying advanced neural network algorithms on efficient computational architectures to solve complex tasks. Specifically, we seek submissions reporting concrete measured results in terms of energy/power/throughput performance with at or near state-of-the-art accuracy on all manner of datasets. We also welcome research exploring the tradeoffs of accuracy and energy at the algorithmic and architectural levels as well as optimization of algorithms for energy efficiency on such platforms. In all cases, papers that report concrete measured power/energy results will be given higher consideration than simulated results.

Scope and Topics

This session seeks contribution on the latest developments and emerging research for all aspects of energy-efficient neural networks. Specific topics for the session include, but are not limited to: Topics:

  • measured energy/power/throughput performance with at or near state-of-the-art accuracy
  • energy efficient architectures running NN algorithms
  • optimization of NN algorithms for energy efficiency
  • energy-accuracy performance tradeoffs for NN algorithms
  • high precision NNs using low precision weights and parameters
  • precision-energy-accuracy performance tradeoffs
  • approximate computing for energy-efficiency
  • low-power embedded applications including algorithm and architecture
  • design of energy-efficient algorithms for specialized architectures
  • energy efficiency of NN algorithms on standard architectures
  • the role of hierarchy in efficient representation
  • the role of attention in efficient computation
Where algorithms and architectures may include, but are not limited to:

Algorithms:
  • Neural Networks (NNs)
  • Convolutional Neural Networks (CNNs)
  • Deep Neural Networks (DNNs)
  • Recurrent Neural Networks (RNNs)
  • Long Short Term Memories (LSTMs)
  • Machine learning algorithms
  • Sparse-distributed representations
  • Sparse encoding
  • Approximate computing algorithms
  • Algorithms for computing with low precision weights
  • Spatio-temporal dynamics/features
  • Hierarchical network topologies
  • Models of attention/saliency
Architectures:
  • Neuromorphic hardware
  • CPUs
  • GPUs
  • Embedded processors
  • FPGAs
  • Coarse-Grained Reconfigurable Arrays (CGRA)
  • DSPs
Applications:
  • Vision, speech, natural language,
  • Cognition, perception, recognition
  • Robotics
  • Decision making under uncertainty

IJCNN-39 Hybrid Neural Intelligent Systems

Organized by Patricia Melin

This Special Session is being organized as one of the main activities of the Task Force on Hybrid Intelligent Systems of the NNTC and will consist of papers that integrate different Soft Computing (SC) methodologies for the development of hybrid neural intelligent systems for modeling, simulation and control of non-linear dynamical systems. The goal of the special session is to promote research on hybrid neural systems all over the world, and researchers working on this area are welcome to submit their papers.

SC methodologies at the moment include (at least) Neural Networks, Fuzzy Logic, Genetic Algorithms and Chaos Theory. Each of these methodologies has advantages and disadvantages and many problems have been solved, by using one of these methodologies. However, many real-world complex industrial problems require the integration of several of these methodologies to really achieve the efficiency and accuracy needed in practice.

Scope and Topics

This session will include papers dealing with methods for integrating the different SC methodologies and neural networks in solving real-world problems. The Special Session will consider applications on the following areas:

  • Robotic Dynamic Systems
  • Non-linear Plants
  • Manufacturing Systems
  • Pattern Recognition
  • Time Series Prediction
Hybrid models offer advantages when a prudent combination of methods is performed and in this case can be a powerful tool in solving complex problems. Research in the area of hybrid neural models is very important and for this reason this session would be of great interest to IJCNN 2016 participants. In fact, previous versions of this session have been organized with great success at WCCI 2008, WCCI 2010, IJCNN 2011, IJCNN 2013 and IJCNN 2014 with over 10 accepted submissions in each case.

IJCNN-40 Computational Intelligence for Safer and Smarter Societies

Organized by Seiichi Ozawa, Cesare Alippi, Sung-Bae Cho and Masahide Nakamura

Thanks to recent advancements in ICT and sensor/actuator technologies, a new generation of devices is made available to constitute smart-grids/homes/buildings/environments and cyber-physical systems. Information coming from the field and feedback actions can be utilized to enriching and improve our lives. In the smart home framework, for example, various sensors are monitoring human behaviors and their health conditions, electrical consumption of appliances and power generation from solar panels; all these systems/applications are then controlled efficiently based on sensed information and inferred human intention. “Smart” technologies received recently a great interest, with smartification not limited to personal houses only. In fact, technologies, intelligent processing and actions can be extended to all kinds of private/public services and our living environments in the large. Smartification would bring us a new form of Society, a “Smart Society”, which provides not only comfortable but also safe environments both in the physical and the cyber worlds. Therefore, ensuring safety and security is also a major issue that a Smart Society has to address and provide. Smart technologies will also enable the design of applications protecting us from risks, crime and hazards, such as natural disasters, accidents, terrorism, cyber-attacks to leaking privacy via phishing.

Scope and Topics

The purpose of this special session is to share new research about computational intelligence and its fundamental role in building, operating, and driving a safer and smarter society. As such, we welcome high-quality and unpublished papers focusing on how computational intelligence techniques can shape smart systems, methodologies and their applications. The topics of interest of this special session include, but are not limited to:

  • Smart Homes and Smart Cities
  • Healthcare and Smart Elderly Care
  • Smart Grids
  • Smart Infrastructures
  • Smart and Precise Agriculture
  • Smart Transportation Systems
  • Hazard and Accident Detection/Prediction
  • Homeland Security
  • Cyber-security and Privacy Protection

IJCNN-41 Incremental Machine Learning: Methods and Applications

Organized by Seiichi Ozawa, Nistor Grozavu and Nicoleta Rogovschi

Thanks to recent advancement of ICT and sensor technologies, data are continuously generated by various kinds of information sources (e.g., SNS, e-mail, POS systems, surveillance cameras, etc.) and such a sequence of data is often called “data stream”. To learn from a data stream in real time, learning must be carried out incrementally in one pass with incoming data. Therefore, fast incremental machine learning algorithms are solicited to learn useful features and classifiers, and predictor. However, learning data streams in one pass is a challenging problem when we have to deal with high-dimensional large-scale data, unbalanced data or data with serious noise/outliers. This Special Session aims for brining and sharing new ideas on new types of incremental learning algorithms under stationary and non-stationary environments.

Scope and Topics

A wide range of incremental online learning methods and applications is covered in this special session, including but not limited to the followings: Theory:

  • Incremental Supervised Learning
  • Incremental Unsupervised Learning
  • Online Learning
  • Online Feature Selection
  • Clustering data stream
  • Distributed Clustering
  • Consensus Clustering
  • Incremental Probabilistical Models
  • Active Learning
    Application:
  • Incremental learning for data mining
  • Incremental learning for computer vision and speech processing
  • Incremental learning for robotics
  • Incremental learning for web

IJCNN-42 Theoretical Foundations of Deep Learning Models and Algorithms

Organized by Alessandro Sperduti, Jose C. Principe and Plamen Angelov

Deep learning models and techniques are becoming more and more the computational tool of choice when facing difficult applicative problems, such as speech and image understanding. The reason for this huge interest in deep learning is due to the fact that their adoption leads to human (and, in some cases, super-human) performances. These successes, however, have been mainly obtained on empirical basis, often thanks to the computational power provided by parallel computer facilities such as GPUs or CPU clusters. Although some recent works have addressed deep learning from a theoretical perspective, still there is a limited understanding of why deep architectures work so well and on how to design computationally efficient and effective training algorithms.

Scope and Topics

This special session aims to gather together leading scientists in deep learning and related areas within computational intelligence, neuroscience, machine learning, artificial intelligence, mathematics, and statistics, interested in all aspects of deep architectures and deep learning, with a particular emphasis on understanding fundamental principles. Topics of interest to the special session include, but are not limited to:

  • Theoretical results on representation and learning in natural or artificial deep architectures
  • Theoretical and/or empirical analysis of specific natural or artificial deep architectures or algorithms
  • Innovative deep architectures/algorithms for data representation and analysis, including both supervised methods like deep convolution networks and unsupervised ones like stacked auto-encoders and deep Boltzmann machines
  • Design and/or analysis of recurrent and recursive architectures for processing of sequences and more general data structures
  • Applications of deep learning in data representation and analysis, including recognition, understanding, detection, segmentation, retrieval, restoration, super-resolution, and compression
  • Deep learning algorithms that efficiently handle large-scale data

IJCNN-43 Neural Network Transfer Learning for the Recognition of Human Behavior and Affect

Organized by Friedhelm Schwenker and Stefan Scherer

The proposed special session focuses on neural network-based transfer learning and knowledge adaptation for pattern recognition problems in human-computer interaction scenarios. Of particular interest for the special session is the classification of human behavior patterns and affect.

Scope and Topics

The special session’s topics include but are not limited to:

  • Learning from multiple sources
  • Deep learning architectures
  • Multi instance learning
  • Multi label learning
  • Learning from unlabeled data
  • Learning from partially labeled data
  • Affective Computing
  • Human Behavior Analysis
  • Intelligent interaction, Assistive systems, Companion systems

IJCNN-44 Deep Reinforcement Learning (DRL)

Organized by Abdulrahman Altahhan, Vasile Palade, Junyu Dong, Xinghui Dong, Hui Yu and Mohamed Cheriet

Deep Learning has been under the focus of neural network research and industrial communities due to its proven ability to scale well into difficult problems and due to its performance breakthroughs over other architectural and learning techniques in important benchmarking problems. This was mainly in the form of improved data representation of supervised learning tasks. Reinforcement learning (RL) is considered the model of choice for problems that involve learning from interaction, where the target is to optimize a long term control strategy or to learn to formulate an optimal policy. Typically these applications involve processing a stream of data coming from different sources that varies between central massive database to pervasive smart sensors (such as the one that is commonly used by a diabetic person or smart home thermostat).

RL do not lend itself naturally to deep learning and currently there is no uniformed approach to combine deep learning with reinforcement learning despite good attempts. Important questions still open; for example how to make the state-action learning process deep? How to make the architecture of an RL system pertains to deep learning without compromising the interactivity of the system? Although recently there have been important advances in dealing with these issues, they are still scattered with no overarching framework that is well defined in a natural way.

This special session will provide a unique platform for researchers from the Deep Learning and Reinforcement Learning communities to share their research experience towards a uniformed Deep Reinforcement Learning (DRL) framework in order to allow this important interdisciplinary branch to take-off on solid grounds. It will concentrate on the potential benefits of the different approaches to combine RL and DL. It aims at bringing more focus to the potential of infusing reinforcement learning framework with deep learning capabilities that could allow it to deal more efficiently with current learning application including but not restricted to online streamed data processing that involves actions.

Scope and Topics

Topics of interest include, but are not limited to the following:

  • Novel DRL Algorithms
  • Novel DRL Neural Architecture
  • Adaptation of existing RL Techniques for Deep Learning
  • Optimization and convergence proofs for DRL algorithms
  • Deeply Hierarchical RL
  • DRL architecture and algorithms for Control
  • DRL architecture and algorithms for Robotics
  • DRL architecture and algorithms for Time Series
  • DRL architecture and algorithms for Big Streamed Data Processing
  • DRL architecture and algorithms for Governmental policy optimization
  • DRL Application in general

IJCNN-45 Neuro-Inspired Computing with Beyond-CMOS Technology

Organized by Saibal Mukhopadhyay and Kaushik Roy

The neuro-inspired and non-Boolean algorithms are emerging as a strong candidate for future computing platforms. The application domain ranges from smart sensor to accelerators in mobile platforms to high-performance systems. The digital CMOS‐based hardware realization of these platforms demonstrates limited energy‐efficiency. Consequently, there is a strong need and interest in exploring beyond‐CMOS technologies for hardware platforms for neuro‐inspired algorithms. The potential technologies include alternative field‐effect‐transistors like Tunneling transistors; emerging memory devices like Resistive RAMs, memristors; and non‐charge-based devices like Spintronics. The focus on this special session is to highlight the recent advancements of application of these emerging devices to various types of neuro‐inspired platforms including associative memory, different neural networks like Cellular Neural Network, Spiking Neural Network, and oscillatory computing. The presented papers will highlight algorithm, architecture, and technology co‐design approaches, and comparative analysis with digital CMOS‐based implementations. The proposed special session will provide analysisforum for fellow researchers in this exciting cross‐disciplinary field.

Scope and Topics

The topics of the special session include, but are not limited to:

  • Digital‐CMOS implementation of neuro­‐inspired algorithms
  • Neuro‐inspired platform with analog CMOS
  • Neuromorphic computing with non­‐CMOS field­‐effect­‐transistor
  • Spintronics‐based neuro­‐inspired computin
  • Neuromorphic platforms with emerging memor
  • Oscillatory computing with non­‐CMOS device

IJCNN-46 Dynamics and Design of Neural Networks and Applications in Industry

Organized by Zhanshan Wang, Guotao Hui and Tieshan Li

An artificial neural network is an information processing paradigm that is inspired by the way biological nervous systems, such as the brain, process information. The key element of this paradigm is the novel structure of the information processing system. With their remarkable ability to derive meaning from complicated or imprecise data, neural networks can be provided to extract patterns and detect trends that are too complex to be noticed by either humans or other computer techniques. Traditional neural network schemes have been applied to several practical applications, such as continuous stirred tank reactor, manipulator arm, image recognition, image encryption, etc. Nonetheless, numerous other real-time applications, such as secure communication, cordless electric cars, machine tools, memory storage, power grid, contain complicated historical data, complex mechanism characteristics and so on. So this special session aims at disseminating the latest interdisciplinary research on the theory and application of neural networks.

The purpose of this special session is to provide an opportunity for scientists, engineers, and practitioners to propose their latest theoretical and technological achievements in the analysis of dynamics and design of neural networks and application in industry. Besides, this special session aims at bringing together researchers in neural networks (NNs) and related areas to brainstorm about new solutions and directions.

Scope and Topics

Authors are invited to submit their original work on neural networks and related fields. The potential topics include, but are not limited to:

  • Reinforcement learning
  • NN-based fault-tolerant control
  • Qualitative analysis and control of neural networks
  • Neural Network Optimization
  • Solving optimization problems arising in economy, society and engineering via neural networks
  • Modeling and analysis of high-order neural networks
  • Predicting Physical Health
  • Data mining based on neural network
  • Modeling, analysis and application of NN in clean energy
  • Modeling, analysis and application of stochastic neural networks and fuzzy neural networks

IJCNN-47 Computational Intelligence Paradigms For Space Weather Prediction

Organized by Mahboobeh Parsapoor and John Brooke

Space weather can be defined as changes in the solar system that causes solar wind and geomagnetic storms and influence magnetosphere, ionosphere and thermosphere that in turn causes damage both Ground-based and space-based communication systems and human life.

Scope and Topics

The main goal of this Special Session is to examine Computational Intelligence Paradigms for predicting solar activity and geomagnetic storms to develop alert system based on them. Other aims of this session are to:

  • Evaluate the current Brain-inspired connectionist approaches on space weather prediction.
  • Provide an opportunity for different researchers to share their ideas.
  • Present new advances in the above approaches.
The submitted papers can utilize one of these methods:
  • neural networks
  • neuro-fuzzy methods
  • deep learning
  • emotion-based data-driven methods
  • hierarchical temporal memory
to predict one of the aspects of space weather phenomena.

IJCNN-48 Affective Brain-Computer Interaction

Organized by Bao-Liang Lu, Thierry Pun, Milos R. Popovic and Hiroshi Yokoi

Over the last decade, there has been a rising tendency in affective Brain-Computer Interaction (aBCI) research to enhance Brain-Computer Interaction systems with the ability to detect, process, and respond to users emotional states. Besides logical intelligence, the introduction of emotional intelligence into BCI to create aBCI has received increasing interest from interdisciplinary research fields including psychology, neuroscience, computer science, and computational intelligence. In this new domain of affective sciences, aBCI aims to narrow the communication gap between the highly emotional human and the BCI systems by developing computational systems that recognize and respond to human emotions. Various applications of aBCI systems have been proposed such as workload monitoring, driving fatigue detection, implicit affective tagging, and game adaptation.

With the fast development of embedded systems and wearable technology, it is now conceivable to port aBCI systems from laboratory to real-world environments. Various advanced dry electrodes and embedded systems including some commercial products are developed to handle the wearability, portability, and practical use of these systems in real world applications. aBCI includes affective sensing, emotion detection and feedback from brain signals and other physiological activity, which extends the concept of conventional BCI.

Although significant advances have been made and many applications have been proposed, the problem of detecting, modeling and regulating emotions in aBCI systems remains complex and largely unexplored. There exist many critical challenges in aBCI systems. How can we deal with artifacts and noises in uncontrolled real-world environments? How can machines respond to the recognized affective states and bring users to a desired affective state? How can we elicit and measure emotions in social setting? How can we develop adaptive aBCI systems that address individual differences and changing environments? How can we introduce contextual information to aBCI? What are the neural patterns or signatures for different emotional states and how is the stability of computational models over time? The goal of this special session is to connect researchers from related fields to discuss the state-of-the-art progress and enhance inter-disciplinary collaborations in aBCI. We are soliciting original contributions for addressing the above research questions.

Scope and Topics

Topics of interest include but are not limited to:

  • Emotional intelligent theory,methods and technology for aBCI
  • Multimodal deep learning for aBCI through combining neurophysiological and peripheral physiological signals
  • Thansfer learning for improving performance of aBCI
  • Feature extraction and selection methods for aBCI
  • Evaluation methods for aBCI
  • Wearable technology for aBCI
  • Affective sensing using neurophysiological signals
  • Affective neural-feedback and regulation
  • Emotion elicitation and database development
  • Applications of aBCI to social interactions
  • aBCI for neural rehabilitation such as motor function
  • aBCI for diagnosis of mental diseases such as ADHD and autism

IJCNN-49 Machine Learning Methods Robust to Large Outliers

Organized by Badong Chen and Lei Sun

This special session focuses mainly on various machine learning methods robust to large outliers (or impulsive noises).

Scope and Topics

Topics of interest include:

  • Machine learning methods (supervised or unsupervised) that are insensitive to large outliers
  • Robust machine learning for dealing with heavy-tailed non-Gaussian data
  • Robust methods for outliers detection
  • Robust and sparsity-aware learning
  • Online learning in abruptly changing environments
  • Applications of machine learning to real data with outliers

IJCNN-50 Computational Intelligence for Personal Health

Organized by Alfredo Vellido, José D. Martín and Paulo J.G. Lisboa

Personal health is widely seen as the future of healthcare, with a focus on the 4Ps: Prediction, Prevention, Personalisation and Participatory healthcare.  This is distinct from pharmacogenomics or stratified therapies, but focuses instead on tracking our health – rather than illness – using wearable sensors and other home/wifi sensors to measure our physical state over time. This is a data-rich context in which Computational Intelligence (CI) can provide a wide range of tools for health-related knowledge extraction. Trends can then be identified which are used for a range of purposes from motivation for regular exercising to prevention and early screening for unexpected deterioration. Where individuals suffer from chronic conditions, this approach can help to target the right intervention at the right time.  This topic has links with health vaults, with avatars, as well as supporting care for the specific groups such as the elderly. This is a topic in which data integration is of key importance and different types of data can be brought in to provide context, ranging from demographics to social media.

Scope and Topics

This workshop will focus on methodologies from the fields of Machine Learning (ML) and, in the broadest sense, CI, as well as on prototype applications targeting this emergent area of health care. Suitable topics would include, but are not limited to:


  • Solutions for heterogeneous data integration in personal health applications.
  • Prototype applications of data analysis for personal health.
  • ML and CI-based application in personal health.
  • Intelligent analysis of wearable sensor data in health monitoring applications. 
  • Intelligent analysis of health vaults data.
  • Natural Language Processing for knowledge extraction from doctor’s prescriptions and/or patient’s feedback using mobile and tablet apps.

IJCNN-51 Advanced Methods in Optimization and Machine Learning for Multimedia Computing

Organized by Yiu-Ming Cheung, Yang Liu Yuping Wang and Ping Guo

Recent advances in storage, hardware, information technology, communication, and networking have resulted in a large amount of multimedia data. This has powered the demand to extract useful and actionable insights from such data in an automatic, reliable and scalable way. Machine learning, which aims to construct algorithms that can learn from and make predictions on data intelligently, has attracted increasing attention in the recent years and has been successfully applied to many multimedia computing tasks, such as image processing, face recognition, video surveillance, document summarization, etc. Since a lot of machine learning algorithms formulate the learning tasks as linear, quadratic or semi-definite mathematical programming problems, optimization becomes a crucial tool and plays a key role in machine learning and multimedia data analysis tasks. On the other hand, machine learning and the applications in multimedia computing are not simply the consumers of optimization technology but a rapidly evolving interdisciplinary research field that is itself promoting new optimization ideas, models, and solutions.

Scope and Topics

This special session "Advanced Methods in Optimization and Machine Learning for Multimedia Computing" aims to provide a platform for academics and industry-related researchers in the areas of applied mathematics, machine learning, artificial intelligence, pattern recognition, data mining, multimedia processing, and big data to exchange ideas and explore traditional and new areas in optimization and machine learning as well as their applications in multimedia computing. The topics of the special session include, but are not limited to: 


  • Approximation algorithms
  • Cloud-based multimedia computing
  • Clustering and graph-partitioning for multimedia computing
  • Cross-media learning
  • Distributed/parallel optimization algorithms in machine learning
  • EM algorithm and alternating optimization
  • Extreme learning machines for multimedia computing
  • Feature and subspace selection for multimedia data abstractions
  • Graph-based learning for multimedia networks
  • High-dimensional data visualization
  • Human/crowd behavior analysis via machine learning
  • Implementation issues of optimization and learning in multimedia computing
  • Learning complex social networks
  • Learning for imbalanced multimedia data
  • Learning mechanisms of visual computing
  • Learning on brain-imaging data
  • Learning video content from unmanned aerial vehicle
  • Media content security
  • Metrics and methods to evaluate multimedia quality of experience
  • Mobile multimedia computing
  • Multimedia search and retrieval
  • Multi-objective optimization and many-objective optimization
  • Nonconvex optimization and numerical methods in machine learning
  • Optimization and machine learning in crowdsourcing
  • Optimization for deep models
  • Optimization for large-scale multimedia computing
  • Optimization in evolutionary computation
  • Optimization in statistics, statistical/computational tradeoffs
  • Optimization on manifolds, metric spaces
  • Optimization with sparsity constraints
  • Probabilistic models and graphical models for multimedia computing
  • Regularization and generalization in machine learning
  • Supervised/semi-supervised/unsupervised learning for multimedia computing
  • Sequential learning for video and audio data
  • Social media
  • Support vector machines and kernel methods for multimedia computing

IJCNN-52 Approximate Dynamic Programming and Reinforcement Learning for High Dimensional Systems

Organized by Dongbin Zhao, Yuanheng Zhu and Haibo He

In the past few decades, adaptive dynamic programming (ADP) and reinforcement learning (RL) have been extensively studied from the aspects of both computational intelligence and control communities. Great achievements have been obtained in solving optimal control and providing intelligent solutions for various problems. Multi-agent ADP/RL is a very hot research topic to solve cooperative control problems among agents with multiple inputs. Recently, the input to the agent is extended to higher dimension, e. g., image. Google Deepmind group proposes promising results for video games based on deep reinforcement learning, with videos or images are as the input, which is a great success and inspiration for ADP/RL research domain. In sum, the ADP/RL for high dimensional systems deserves more and more research attention.

Scope and Topics

The aim of this special session is to call for the most advanced research and state-of-the-art works in the field of ADP/RL in high dimensional systems. It is expected to provide a platform for international researchers to exchange ideas and to present their latest research in the relevant topics. All the original papers related to ADP and RL are welcome. Specific topics of interest include but are not limited to:


  • Theoretical foundation of ADP in convergence, stability, robustness, and etc. ;
  • Multi-agent adaptive dynamic programming;
  • Hierarchical reinforcement learning;
  • Deep reinforcement learning;
  • Deep adaptive dynamic programming;
  • Combination of ADP and RL with recurrent neural network;
  • Data-driven learning and control;
  • Applications in realistic and complicated systems.

FIND US ON

line line

FUZZ-IEEE Special Sessions


FUZZ-IEEE 2016 Special Sessions

FUZZ-IEEE-01 Uncertainty Theory and Its Application

Organized by Zutong Wang, Baoding Liu, Dan Ralescu and Jiansheng Guo

In order to deal with indeterminacy mathematically, two axiomatic systems have been founded, namely, probability theory and uncertainty theory. When no samples are available to estimate a probability distribution, we have to invite some domain experts to evaluate the belief degree that each event will happen. In order to rationally deal with personal belief degrees, uncertainty theory was founded in 2007 and subsequently studied by many researchers. Nowadays, uncertainty theory has become a branch of mathematics.

Scope and Topics

The goal of this special session is to provide an excellent forum for the discussion of the latest theoretical advances and practical applications in this exciting research field, to help foster the understanding, development, and practice of uncertainty theory for solving problems in economy, engineering, management and the social sciences. We invite the submission of high-quality, original and unpublished papers in this area. The topics of interest include, but are not limited to:

  • Uncertainty theory
  • statistics
  • Uncertain programming
  • Uncertain risk analysis
  • Uncertain set
  • Uncertain logic
  • Uncertain inference
  • Uncertain process
  • Uncertain calculus
  • Uncertain differential equation
  • Other real-world applications

FUZZ-IEEE-02 Belief Function Theory and Its Applications (Cancelled)

Organized by Yafei Song, Zhun-Ga Liu and Xiaodan Wang

Since its inception, belief function theory, also known as Dempster-Shafer theory or evidence theory, has received growing attention in many fields of applications such as finance, technology, biomedicine, etc, despite of its incompetence in combining belief functions with high conflict. To remove the roadblocks in the development of belief function theory, many improvements have been made subsequently, e.g., the present of transferable belief model (TBM) and Dezert-Smarandache theory (DSmT). At present, more and more researchers are dedicated to studying belief function theory from different views for further exploration and better exploitation.

Scope and Topics

This special session is intended to provide the latest advances of belief function theory, the relationship between belief function theory and other theories such as probability theory, possibility theory, rough set theory, and fuzzy set theory, the fusion of imperfect information in the united framework of random sets theory, together with their applications in artificial intelligence, to enhance the development of belief function theory for solving problems in engineering. We invite original submissions of high quality in this area. The topics of interest include, but are not limited to:

  • Belief function theory
  • Evidence accumulation
  • Temporal information fusion
  • Conflict management
  • Identification fusion
  • Knowledge discovery
  • Data mining
  • Fuzzy sets
  • Rough sets
  • Random sets
  • Applications of artificial intelligence

FUZZ-IEEE-03 Recent Trends in Many-Valued Logic and Fuzziness

Organized by Stefano Aguzzoli, Pietro Codara and Diego Valota

Many-valued logics have constituted for several decades key conceptual tools for the formal description and management of fuzzy, vague and uncertain information. In the last few years, the study of these logical systems has seen a bloom of new research related to the most diverse areas of mathematics and applied sciences. Relevant recent developments in this field are connected to the natural semantics of non-classical events. A nonclassical event is described by a formula in the language of a given manyvalued logic. A satisfying semantics for such events must account for their different aspects, in particular the "ontic" aspect, related to their vague nature, and the "epistemic" aspect, related to our ignorance, or approximate knowledge about them. The combination in a unique conceptual framework of the logic and the probability of a class of non-classical events, usually reached through the algebraic semantics and their topological or combinatorial dualities, provides both the theoreticians and the applicationoriented scholars with powerful tools to deal with this kind of events. This special session is devoted to the most recent development in the realm of many-valued logics, with particular emphasis on theoretical advances related to algebraic or alternative semantics, combinatorial aspects, topological and categorical methods, proof theory and game theory, manyvalued computation. In particular, results directed towards a better understanding of the natural semantics of non-classical events will be appreciated. Further, a special attention is also given to connections and synergies between many-valued logics and other different formal approaches to vague and approximate reasoning, such as Rough Sets, Formal Concept Analysis and Relational Methods.

Scope and Topics

A partial list of topics is the following:

  • Algebraic semantics of many-valued logics
  • Applications of many-valued logics to Formal Concept Analysis and Relational Methods
  • Applications of many-valued logics to Fuzzy Sets and to Rough Sets
  • Combinatorial or topological dualities
  • Computational complexity of many-valued logics
  • Many-valued computational models
  • Modal logic approaches to probability and uncertainty in many-valued logics
  • Natural and alternative semantics for many-valued logics
  • Proof theory for many-valued logics
  • Representation theory
  • Subjective probability approaches to many-valued logics and nonclassical events

FUZZ-IEEE-04 Fuzzy Interpolation

Organized by Qiang Shen, Laszlo Koczy Shyi-Ming Chen and Ying Li

Fuzzy interpolation provides a flexible means to perform reasoning in the presence of insufficient knowledge that is represented as sparse fuzzy rule bases. It enables approximate inferences to be carried out from a rule base that does not cover a given observation. Fuzzy interpolation also provides a way to simplify complex systems models and/or the process of fuzzy rule generation. It allows the reduction of the number of rules needed, thereby speeding up parameter optimisation and runtime efficiency.

Scope and Topics

The aim of this special session is to provide a forum:

  • To disseminate and discuss recent and significant research efforts in the development of fuzzy interpolation and related techniques,
  • To promote both theoretical and practical applications of fuzzy interpolation, and
  • To foster integration of fuzzy interpolation with other computational intelligence techniques.
The topics of this special session will include but are not limited to:
  • Fuzzy interpolation
  • Fuzzy extrapolation
  • Fuzzy interpolative learning
  • Fuzzy systems simplification
  • Fuzzy set transformation
  • Fuzzy set representation
  • Fuzzy interpolation application
  • Fuzzy function approximation
  • Hybrid fuzzy interpolation systems
  • Comparative studies of interpolation methods

FUZZ-IEEE-05 Inter-Relation Between Interval and Fuzzy Techniques (Cancelled)

Organized by Vladik Kreinovich, Hung T. Nguyen and Juan Carlos Figueroa Garcia

The relation between fuzzy and interval techniques is well known; e.g., due to the fact that a fuzzy number can be represented as a nested family of intervals (alpha-cuts), level-by-level interval techniques are often used to process fuzzy data.

At present, researchers in fuzzy data processing mainly used interval techniques originally designed for non-fuzzy applications, techniques which are often taken from textbooks and are, therefore, already outperformed by more recent and more efficient methods.

One of the main objectives of the proposed special session is to make the fuzzy community at-large better acquainted with the latest, most efficient interval techniques, especially with techniques specifically developed for solving fuzzy-related problems.

Another objective is to combine fuzzy and interval techniques, so that we will be able to use the combined techniques in (frequent) practical situations where both types of uncertainty are present: for example, when some quantities are known with interval uncertainty (e.g., coming from measurements), while other quantities are known with fuzzy uncertainty (coming from expert estimates).

Scope and Topics

The topics of this special session will include but are not limited to:

  • interval computations
  • interval uncertainty
  • interval-valued fuzzy sets

FUZZ-IEEE-06 Software for Soft Computing

Organized by Jesús Alcalá-Fdez and José M. Alonso

The term Soft Computing is usually used in reference to a family of several preexisting techniques (Fuzzy Logic, Neuro-computing, Probabilistic Reasoning, Evolutionary Computation, etc.) able to work in a cooperative way, taking profit from the main advantages of each individual technique, in order to solve lots of complex real-world problems for which other techniques are not well suited.

In the last few years, many software tools have been developed for Soft Computing. Although a lot of them are commercially distributed, unfortunately only a few tools are available as open source software. Please, notice that such open tools have recently reached a high level of development. As a result, they are ready to play an important role for industry and academia research.

Scope and Topics

The aim of this session is to provide a forum to disseminate and discuss Software for Soft Computing, with special attention to Fuzzy Systems Software. We want to offer an opportunity for researchers and practitioners to identify new promising research directions in this area. Potential topics of interest include but are not limited to

  • Data preprocessing
  • Data mining and evolutionary knowledge extraction
  • Modeling, Control, and Optimization
  • System validation, verification, and exploratory analysis
  • Knowledge extraction and linguistic/graphical representation
  • Visualization of results
  • Languages for soft computing software
  • Interoperability
  • High Performance Computing (Map-Reduce, GPGPU, etc.)
  • Applications

FUZZ-IEEE-07 Recent Advances and New Challenges in Evolving Fuzzy Systems

Organized by Plamen Angelov, Fernando Gomide, Edwin Lughofer and Igor Skrjanc

Evolving systems are modular systems that simultaneously develop their structure, functionality, and parameters in a continuous, self-organized, one pass adaptive way from data streams.

During the last 12-15 years, the concept of Evolving Fuzzy Systems (EFS) established as a useful and necessary methodology to address the problems of imprecision, incremental learning, adaptation and evolution of fuzzy Systems in dynamic environments and during on-line/real-time operation modes. EFS are able to automatically and autonomously adapt themselves to new operating conditions and system states and hence guarantee a high process safety, especially in case of highly dynamic and time-variant systems. This is especially necessary when precise and sufficient training data is not available (e.g., because of high costs for data collection or annotation) in order to set up models which cover the whole range of possible system states. Another major topic which can be addressed with EFS are the building of models from huge massive stream data or even from Big Data, and to serve as dynamically adaptable knowledge base within enriched human-machine interaction applications (learning and teaching).

Scope and Topics

The goal of the special session is to provide a broad picture of the recent developments and to explore further (open) research challenges in one or several specific research topics mentioned below.

  • Novel adaptive, incremental methods in evolving fuzzy modeling tasks:
    • Evolving fuzzy classifiers
    • Evolving Takagi-Sugeno-Kang type fuzzy systems
    • Evolving neuro-fuzzy approaches
    • Evolving type-2 fuzzy systems and related architectures
    • Evolving modeling and control systems
    • Data stream fuzzy clustering
    • Adaptive fuzzy pattern recognition
    • Adaptive fuzzy regression and correlation techniques
  • Enhanced Issues in dynamic fuzzy methods:
    • Issues on robustness, stability and process-safety in evolving fuzzy systems
    • Evolving techniques to address concept drift and shift
    • Evolving fuzzy models in soft sensing
    • On-line techniques to deal with model uncertainty and interpretability issues
    • Active and semi-supervised learning with fuzzy concepts
    • On-line dimensionality reduction and feature selection
    • Evolving granular modeling and control
    • Towards plug-and-play capability
  • Real-World applications of evolving fuzzy systems in:
    • On-line system identification
    • On-line fault detection and decision support diagnosis
    • Data stream mining and adaptive knowledge discovery
    • Database and web mining
    • Control and decision support systems
    • Image classification and visual Inspection
    • Automation and robotics
    • Control systems
    • Data stream mining and adaptive knowledge discovery
    • Forecasting in financial domains and time-series prediction

FUZZ-IEEE-08 Advances to Type-2 Fuzzy Logic Control

Organized by Tufan Kumbasar and Hao Ying

Type-2 fuzzy logic control is a technology which takes the fundamental concepts in control from type-1 fuzzy logic and expands upon them in order to deal with higher levels of uncertainty presented in many real-world control problems. A variety of control application areas have been addressed with type-2 fuzzy logic, from the control in steel production plants to the control of marine diesel engines and robotic control. For some engineering applications, there is evidence that type-2 fuzzy logic can provide benefits over both traditional forms of control as well as type-1 fuzzy logic. It is the aim of this special session to attract a comprehensive selection of high quality current research in this area of type-2 control, motivating further collaboration and providing a platform for the discussion on future directions of type- 2 fuzzy logic control by active researchers in the field.

Scope and Topics

This special session will address advances in interval type-2 as well as general type-2 fuzzy logic control, including different types of type-2 fuzzy logic control such as the PID type, model-based, neuro-fuzzy and TSK-based. As such, the session aims to provide both an overview of the current research as well as a window into the future of type-2 fuzzy logic control. Topics include, but are not limited to:

  • Interval Type-2 Fuzzy Logic Control
  • General Type-2 Fuzzy Logic Control
  • Type-2 TSK Fuzzy Logic Control
  • PID type Type-2 Fuzzy Logic Control
  • Model-Based Type-2 Fuzzy Logic Control
  • Adaptive / Self-Tuning Type-2 Fuzzy Control
  • Neuro-Fuzzy Type-2 Control
  • Applications of Type-2 Fuzzy Controllers

FUZZ-IEEE-09 Simulation Modeling and Fuzzy Logic (Cancelled)

Organized by Hamidreza Izadbakhsh, Marzieh Zarinbal and Amir Zarinbal

Simulation modeling is broad collection of methods mainly used to imitate the behaviors of real processes or real systems over time. These methods have been applied in many areas from operational level to tactical level and strategic level. Major approaches in simulation modeling could be classified into three categories, discrete-event, system dynamics, and agent based. While, discrete-event modeling is used in operation and tactical level, system dynamics is mainly used in strategic level, and agent based modeling is being used in all levels.

However, there are many situations, in which a real system could not be modeled using traditional simulation methods. The parameters of the system are uncertain, the functions are vague, and the distributions are imprecise. In these situations, fuzzy logic could be applied. In other words, fuzzy simulation methods give more flexibility to handle uncertainties in real situations and have been applied in many areas such as, layout optimization, scheduling, health risks assessment, intelligent transportation system, advanced traffic management systems, pension fund optimization, maintenance planning, portfolio selection, etc.

Scope and Topics

Regarding the interest in this area, this special session looks to gather and discuss the latest theoretical and application achievements in analyzing, designing and optimizing simulation modeling using fuzzy logic. Potential topics include but not limited to:

  • Discrete-event simulation and fuzzy logic
  • System dynamics and fuzzy logic
  • Agent based modeling and fuzzy logic
  • Big data in fuzzy simulation modeling
  • Application of fuzzy simulation modeling in transportation and traffic
  • Application of fuzzy simulation modeling in healthcare
  • Application of fuzzy simulation modeling in manufacturing
  • Application of fuzzy simulation modeling in supply chain management

FUZZ-IEEE-10 Type-2 Fuzzy Sets and Systems Applications (T2-A)

Organized by Christian Wagner and Hani Hagras

Type-2 fuzzy sets and systems are paradigms which seek to realize computationally efficient fuzzy systems with the ability to give excellent performance in the face of highly uncertain conditions. Specifically, type-2 fuzzy sets provide a framework for the comprehensive capturing and modelling of uncertain data, which, together with approaches such as clustering and similarity measures (to name but two) provides strong capability for reasoning about and with uncertain information sources in a variety of contexts and applications. Type-2 fuzzy systems combine the potential of type-2 fuzzy sets with the strengths of rule-based inference in order to provide highly capable inference systems over uncertain data which remain white-box systems (i.e. interpretable).

Scope and Topics

The aim of this special session is to present and focus top quality research in the areas related to the practical aspects and applications of type-2 fuzzy sets and systems. The session will also provide a forum for the academic community and industry to report on recent advances within the type-2 fuzzy sets and systems research. Topics include, but are not limited to:

  • Type-2 Applications
  • Applications including similarity and distance measures for type-2 fuzzy sets
  • Data analysis*
  • Robotics*
  • Decision Making*
  • Clustering and Classification*
  • Modelling*
  • Computing with words*
  • Type-2 Fuzzy Agents
  • Any other application area that employs type-2 fuzzy sets
* using type-2 fuzzy sets and/or fuzzy systems

FUZZ-IEEE-11 Fuzzy and Intelligent Control Systems

Organized by Ching-Chih Tsai

In recent years, a trend has emerged in which techniques of computational intelligence; learning control and automation have been integrated into intelligent control or automation systems on a variety of scales to meet the needs of implementation at the angle of products. Many computational intelligence and learning methods, including fuzzy control, neural networks, fuzzy neural networks, CMAC, genetic algorithm, artificial immune networks, swarm particle techniques, ACO, reinforcement learning and etc., have gained successful applications in many industrial control automation fields. In light of this emerging trend, we propose a special session, called “ fuzzy and intelligent control systems”, at FUZZ-IEEE 2016, in order to promote the advanced theory, practice, and interdisciplinary aspects of integration of computational intelligence and learning control in the area of intelligent control and automation systems. This special session aims to disseminate high quality research results regarding not only the theoretic development in integration of computational intelligence theories and control techniques, but also related effective applications to some new and useful physical systems. In this proposal, particular attention will be paid to highly selected topics about novel fuzzy, intelligent control and learning methods for uncertain systems.

FUZZ-IEEE-12 The Theory of Type-2 Sets and Systems (T2-T)

Organized by Simon Coupland, Robert John and Jonathan Garibaldi

Type-2 fuzzy sets and systems are paradigms which seek to realize computationally efficient fuzzy systems with the ability to give excellent performance in the face of highly uncertain conditions. Specifically, type-2 fuzzy sets provide a framework for the comprehensive capturing and modelling of uncertain data, which, together with approaches such as clustering and similarity measures (to name but two) provides strong capability for reasoning about and with uncertain information sources in a variety of contexts and applications. Type-2 fuzzy systems combine the potential of type-2 fuzzy sets with the strengths of rule-based inference in order to provide highly capable inference systems over uncertain data which remain white-box systems (i.e. interpretable).

Scope and Topics

The aim of this special session is to present and focus top quality research in the areas related to the underlying theory of type-2 fuzzy sets and systems. There are many open and unanswered questions about properties and nature of type-2 fuzzy sets and systems, this session is designed to provide a forum for the academic and industrial communities to report on advances in including, but are not limited to:

  • Representations of type-2 fuzzy sets
  • Approaches to defuzzification
  • Fuzzy operators
  • Fuzzy measures
  • Interpretability
  • Computational complexity
  • Related extensions to type-1

FUZZ-IEEE-13 Fuzzy Set Theory in Computer Vision

Organized by Derek T. Anderson, Chee Seng Chan and James M. Keller

Fuzzy set theory is the subject of intense investigation in fields like control theory, robotics, biomedical engineering, computing with words, knowledge discovery, remote sensing and socioeconomics, to name a few. However, in the area of computer vision, other fields, e.g., machine learning, and communities, e.g., PAMI, ICCV, CVPR, ECCV, NIPS, are arguably state-of-the-art. In particular, the vast majority of top performing techniques on public datasets are steeped in probability theory. Important questions to the fuzzy set community include the following. What is the role of fuzzy set theory in computer vision? Does fuzzy set theory make the most sense and biggest impact in terms of low-, mid- or high-level computer vision? Furthermore, do current performance measures favor machine learning approaches? Last, is there additional benefit that fuzzy set theory brings, and if so, how is it measured?

Scope and Topics

This special session invites new research in fuzzy set theory in computer vision. It is a follow up to the 2013 FUZZ-IEEE workshop View of Computer Vision Research and Challenges for the Fuzzy Set Community and Fuzzy Set Theory in Computer Vision special sessions in 2014 and 2015. In particular, we encourage authors to investigate their research using public datasets and to compare their results to both fuzzy and non-fuzzy methods. Topics of interest include all areas in computer vision and image/video understanding. Example topics include, but are definitely not limited to, the following:

  • Detection and recognition
  • Categorization, classification, indexing and matching
  • 3D-based computer vision
  • Advanced image features and descriptors
  • Motion analysis and tracking
  • Linguistic description and summarization
  • Video: events, activities and surveillance
  • Intelligent change detection
  • Face and gesture
  • Low-level, mid-level and high-level computer vision
  • Data fusion for computer vision
  • Medical and biological image analysis
  • Vision for Robotics

FUZZ-IEEE-14 Fuzzy Systems on Renewable Energy

Organized by Faa-Jeng Lin and Hong-Tzer Yang

Renewable power generation systems in general include wind, photovoltaic (PV), fuel cell and biomass power generation systems. They have been getting more attention recently due to cost competitiveness and environment friendly, as compared to fossil fuel and nuclear power generations. Owing to the relatively higher investment cost of renewable power generation systems, it is important to operate the systems near their maximum power output point, especially for the wind and solar PV generation systems. Thus, maximum power point tracking (MPPT) techniques are often required. Moreover, since the wind and solar PV power resources are intermittent, accurate predictions and modeling of wind speed and solar insolation are necessary, though difficult. Plus, to have a more reliable power supply, renewable power generation systems are usually interconnected with the electrical network. As a result, modeling and controlling the electrical network using smart-grid techniques, such as smart meter, micro-grid, and distribution automations become very important issues. On the other hand, due to the highly nonlinear and time-varying nature with unmodeling dynamics, effective uses of computational intelligence techniques such as fuzzy systems for the controlling and modeling of renewable power generation in a smart-grid system turn out to be very crucial for successful operations of the systems. Hence, topics of interest of the special session on Fuzzy Systems of Renewable Energy would cover the whole range of researches and applications of fuzzy systems in renewable power generations and smart grid systems.

Scope and Topics

Topics of interest include, but are not limited to, the following:

  • Fuzzy modeling of renewable power generation systems
  • Fuzzy control of renewable power generation systems
  • Prediction of renewable energy using fuzzy systems
  • Hybrid systems of computational intelligence techniques in renewable power generation systems
  • Fuzzy energy management systems
  • Fuzzy distribution systems automation
  • Fuzzy power quality, protection and reliability analysis of power system

FUZZ-IEEE-15 Methods and Applications of Fuzzy Cognitive Maps

Organized by Elpiniki I. Papageorgiou, Engin Yesil and Jose Salmeron

Fuzzy Cognitive Map is an extension of cognitive maps for modeling complex causal relationships easily, both qualitatively and quantitatively. As a Soft Computing technique it is used for causal knowledge acquisition and providing causal knowledge reasoning process. FCMs modeling approach resembles human reasoning; it relies on the human expert knowledge for a domain, making associations along generalized relationships between domain descriptors, concepts and conclusions. FCMs can be constructed from raw data as well. FCMs model any real world system as a collection of concepts and causal relation among concepts. They combine fuzzy logic and recurrent neural networks inheriting their main advantages. From an Artificial Intelligence perspective, FCMs are dynamic networks with learning capabilities, whereas more and more data is available to model the problem, the system becomes better at adapting itself and reaching a solution. They gained momentum due to their dynamic characteristics and learning capabilities. These capabilities make them essential for modeling and decision making tasks as they improve the performance of these tasks.

During the past decade, FCMs played a vital role in the applications of diverse scientific areas, such as social and political sciences, engineering, information technology, robotics, expert systems, medicine, education, prediction, environment etc.

Scope and Topics

This special session aims to present highly technical papers describing new FCM models and methodologies addressing any of the following specific topics: theoretical aspects, learning algorithms, innovative applications and FCMs extensions. Topics include, but are not limited to:

  • Modeling Fuzzy Cognitive Maps
  • Approximate Reasoning
  • Knowledge Representation
  • Learning Algorithms for FCMs
  • Evolutionary Fuzzy Cognitive Maps
  • Granular Cognitive maps
  • Rule Based Fuzzy Cognitive Map
  • Fuzzy Cognitive Agents
  • Dynamic Cognitive Networks
  • FCM extensions
  • Fuzzy Grey Cognitive Maps
  • Rough Cognitive Map
  • Intuitionistic Fuzzy Cognitive Maps
  • Interval Fuzzy Cognitive Maps
  • Competitive Fuzzy Cognitive Maps
  • Hybrid FCM-based approaches
  • FCMs for Stakeholders analysis
  • FCMs in Biomedical Engineering
  • FCMs in Pattern Recognition
  • FCMs in Medical Decision Support
  • FCMs in Decision Making and Control Systems
  • FCMs in Business Management
  • FCMs in Engineering
  • FCMs in Agricultural Systems
  • FCMs in Data Mining
  • FCMs in Computer Vision Tasks

FUZZ-IEEE-16 Evolutionary Fuzzy Systems

Organized by Yusuke Nojima, Rafael Alcalá and Hisao Ishibuchi

For more than two decades, evolutionary computation and various meta-heuristics have frequently been used for fuzzy system design under the name of evolutionary fuzzy systems. Their learning and adaptation capabilities enable structure and parameter optimization of fuzzy systems for many kinds of machine learning tasks such as modeling, classification, and rule mining. Their flexible frameworks also enable to handle multiple objectives like accuracy and interpretability maximization and many kinds of data types like imbalanced, missing, and privacy-preserving data sets. The aim of the session is to provide a forum to disseminate and discuss recent and significant research efforts on Evolutionary Fuzzy Systems in order to deal with current challenges on this topic.

Scope and Topics

The session is open to any high quality submission from researchers working at the particular intersection of evolutionary algorithms and fuzzy systems. The topics of this special session are as follows:

  • Evolutionary Learning/Tuning of Fuzzy Rule-Based Systems
  • Evolutionary Selection of Fuzzy Rules
  • Interpretability-Accuracy Tradeoff
  • Multiobjective Evolutionary Fuzzy Systems
  • Evolutionary Fuzzy Neural Networks
  • Evolutionary Fuzzy Clustering
  • Swarm Intelligence for Fuzzy Systems
  • Preprocessing and Postprocessing for Evolutionary Fuzzy Systems
  • Applications of Evolutionary Fuzzy Systems to Real World Problems

FUZZ-IEEE-17 Linguistic Summarization and Description of Data

Organized by Nicolas Marin, Daniel Sanchez, Anna Wilbik and Rui Jorge Almeida

Linguistic summaries and descriptions of data aim to extract and represent knowledge in the form of a collection of natural language sentences. The objective is to obtain a text, as if it was produced by a human expert, describing the most relevant aspects of data for a certain user in a specific context. Automatic generations of data summaries have gained increased relevance with the advent of possibilities to store and acquire data as well as relations between them. In this realm, not only specialized users (e.g. in decision support systems) are interested in this type of approach, but non-specialized users also show interest in receiving understandable information that is supported by data. Linguistic summaries commonly use fuzzy set theory to model linguistic variables and incorporate different forms of imprecision in a collection of natural language sentences. In many approaches they can be considered as quantifier based sentences, hence linguistic summaries constitute a perfect application for new developments in the domain of fuzzy quantifiers. Furthermore, linguistic summaries have been related to fuzzy rule systems.

Linguistic summaries and description of data is related to other research areas such as knowledge discovery in databases and intelligent data analysis, flexible query answering systems for data, human-machine interaction, uncertainty management, heuristics and metaheuristics, natural language generation or processing. More recently, this field has been related to different paradigms, namely the linguistic description of complex phenomena and computing with words paradigms.

The objective of this special session is to provide a forum for researchers, from the above indicated areas, to present recent developments in linguistic summarizes and description of data as well as discuss how these different approaches can complement each other for the task of building such systems.

Scope and Topics

Topics of interest include, but are not restricted to:

  • Protoforms and fuzzy concepts for the linguistic summaries and fuzzy description.
  • Quality assessment of linguistic summaries and fuzzy description.
  • Techniques and algorithms for generating linguistic summaries and descriptions of data.
  • Ontologies for data summarization. Logical approaches for modeling linguistic expressions.
  • Modeling uncertainty for linguistic summaries and fuzzy description.
  • User preference/interest modeling for linguistic summaries and fuzzy description.
  • Applications of linguistic summaries and fuzzy description.
  • Natural language generation for data summarization.
  • Machine Learning applied to data summarization.
  • Linguistic information extraction from visual information Context-awareness in data summarization and description, and natural language generation.

FUZZ-IEEE-18 Handling Uncertainties in Big Data by Fuzzy Systems

Organized by Jie Lu, Chin-Teng Lin, Guangquan Zhang, Farookh Khadeer Hussain, Vahid Behbood, Dianshuang Wu, Mahardhika Pratama and Mohsen Naderpour

The volume, variety, velocity, veracity and value of data and data communication are increasing exponentially. The “Five Vs” are the key features of big data, and also the causes of inherent uncertainties in the representation, processing, and analysis of big data. Also, big data often contains a significant amount of unstructured, uncertain and imprecise data.

Fuzzy sets, logic and systems enable us to efficiently and flexibly handle uncertainties in big data in a transparent way, thus enabling it to better satisfy the needs of real world big data applications and improve the quality of organizational data-based decisions. Successful developments in this area have appeared in many different aspects, such as fuzzy data analysis technique, fuzzy data inference methods and fuzzy machine learning. In particular, the linguistic representation and processing power of fuzzy sets is a unique tool for bridging symbolic intelligence and numerical intelligence gracefully. Hence, fuzzy techniques can help to extend machine learning in big data from the numerical data level to the knowledge rule level. It is therefore instructive and vital to gather current trends and provide a high quality forum for the theoretical research results and practical development of fuzzy techniques in handling uncertainties in big data.

This special session aims to offer a systematic overview of this new field and provides innovative approaches to handle various uncertainty issues in big data presentation, processing and analysing by applying fuzzy sets, fuzzy logic, fuzzy systems, and other computational intelligent techniques.

Scope and Topics

The main topics of this special session include, but are not limited to, the following:

  • Fuzzy rule-based knowledge representation in big data processing
  • Information uncertainty handling in big data processing
  • Unstructured big data visualization
  • Uncertain data presentation and fuzzy knowledge modeling in big data sets
  • Tools and techniques for big data analytics in uncertain environments
  • Computational intelligence methods for big data analytics
  • Techniques to address concept drifts in big data
  • Methods to deal with model uncertainty and interpretability issues in big data processing
  • Feature selection and extraction techniques for big data processing
  • Granular modeling, classification and control
  • Fuzzy clustering, modeling and fuzzy neural networks in big data
  • Evolving and adaptive fuzzy systems in in big data
  • Uncertain data presentation and midelling in data-driven decision support systems
  • Information uncertainty handling in recommender systems
  • Uncertain data presentation and midelling in cloud computing
  • Information uncertainty handling in social network and web services
  • Real world cases of uncertainties in big data

FUZZ-IEEE-19 Interval-Valued Fuzzy Sets: Theory and Applications

Organized by Humberto Bustince , Radko Mesiar, Javier Fernandez and Javier Montero

Since its introduction by Zadeh in 1965 it was clear that fuzzy theory was an extraordinary tool for representing human knowledge. Nevertheless, L. Zadeh himself established in 1973 that sometimes, in decision-making processes, knowledge is better represented by means of some generalizations of fuzzy sets. In the applied field, in particular, the success of the use of fuzzy set theory depends on the choice of the membership function that we make. However, there are applications in which experts do not have precise knowledge of the membership function that should be taken. In these cases, it is appropriate to represent the membership degree of each element to the fuzzy set by means of an interval. From these considerations arises the extension of fuzzy sets called theory of Interval-valued Fuzzy Sets (IVFSs), that is, fuzzy sets such that the membership degree of each element of the fuzzy set is given by a closed subinterval of the interval [0, 1].

The theory of interval-valued fuzzy sets has attracted a lot of interest since its origin and specially in last years, when some applications where the use of intervals have allowed to improve the results of some well-known algorithms in classification or image processing, for instance.

Scope and Topics

This special session will be dedicated to theoretical and practical aspects of interval-valued fuzzy sets. We hope to bring together some of the leading experts in this field, as well as researchers interested in this field, to share their work. In particular, this session covers (but it is not limited to) the following topics:

  • Interval-valued aggregation functions
  • Interval-valued measures and integrals
  • Interval-valued comparison measures
  • Interval-valued fusion functions
  • Interval-valued image processing
  • Interval-valued approximate reasoning
  • Interval-valued classification
  • Interval-valued machine learning
  • Big data with interval-valued fuzzy sets

FUZZ-IEEE-20 Intelligent Medical Science

Organized by Syoji Kobashi, Gerald Schaefer, Hiroharu Kawanaka and Atsushi Inoue

The purpose of this special session is to disseminate and discuss recent and significant research issues on how intelligent methodologies can be used to solve challenging problems related to medical, biomedical, and healthcare fields. This special session will be held under IEEE CIS Task Force of "Fuzzy Logic in Medical Sciences".

Scope and Topics

  • Fuzzy logic-based medical diagnosis control system
  • Fuzzy logic-based biomedical applications
  • Soft computing for biomedical applications
  • Fuzzy logic-based affective computing and psychological evaluations
  • Fuzzy data analysis – bioinformatics, medical informatics, pattern recognition
  • Fuzzy optimization and control Fuzzy machine learning approach to biomedical applications
  • Neuro-fuzzy models for biomedical signal processing
  • Signal processing of MRI, fMRI, EEG, ECG, etc.
  • Smart diagnostic predictions of various diseases
  • Applications in image processing and pattern recognition
  • Fuzzy logic vs. other soft computing approaches
  • Approaches based on neuro-fuzzy, evolutionary neuro-fuzzy, neuro-genetic, genetic fuzzy, fuzzy cognitive map
  • Fuzzy inference systems
  • Assistive robotics
  • Fuzzy temporal representation of knowledge

FUZZ-IEEE-21 Adaptive Fuzzy Control for Nonlinear Systems

Organized by Valentina E. Balas, Tsung-Chih Lin, Rajeeb Dey, Yu-Chen Lin and Seshadhri Srinivasan

The aim of this special session is to present the state-of-the-art results in the area of adaptive intelligent control theory and applications and to get together researchers in this area. Adaptive control is a technique of applying some methods to obtain a model of the process and using this model to design a controller. Especially, fuzzy adaptive control has been an important area of active research. Significant developments have been seen, including theoretical success and practical design. One of the reasons for the rapid growth of fuzzy adaptive control is its ability to control plants with uncertainties during its operation.

Scope and Topics

The papers in this special session present the most advanced techniques and algorithms of adaptive control. These include various robust techniques, performance enhancement techniques, techniques with less a-priori knowledge and nonlinear intelligent adaptive control techniques. This special session aims to provide an opportunity for international researchers to share and review recent advances in the foundations, integration architectures and applications of hybrid and adaptive systems. Topics of interest include, but are not limited to:

  • Fuzzy Self-Organizing Controllers
  • Adaptive Fuzzy Control Design
  • Fuzzy Applications
  • Fuzzy Modeling and Simulation
  • Fuzzy Model Reference Learning Controller
  • Hybrid adaptive fuzzy control
  • Robust adaptive fuzzy control
  • Adaptive fuzzy sliding-mode control
  • Time-Delay Nonlinear Systems
  • Adaptive and learning control theory
  • Adaptive control of processes
  • Data based auto-tuning of the controller
  • Estimation and identification and its application to control design
  • Cooperative Control
  • Hybrid Intelligent Control

FUZZ-IEEE-22 Fuzzy Decision-Making: Consensus and Missing Preferences

Organized by Enrique Herrera-Viedma, Francisco Chiclana, Yucheng Dong and Francisco Javier Cabrerizo

The development of formal mathematical models to support experts in making decisions is of great importance to assure the validity of the actions derived from a decision outcome is theoretically sound. This is of special relevance in decision contexts where the information on the problem at hand is not amenable to be modelled in a quantitative and precise way. Another issue to be addressed is that of inconsistency of information and the dynamic nature of the decision making process itself. This type of decision-making is now being described as decision-making under uncertainty in inconsistent and dynamic environments.

Scope and Topics

This special session aims at gathering researchers with an interest in the research area described above. Specifically, we are interested in contributions towards the development of consensus models for such decision-making problems, as well as formal approaches that are able to support incomplete or missing information.

Contributions to this special session are expected to pay special attention to the rigorous motivation of the approaches put forward and to support all aspects of the models developed with a corresponding theoretical sound framework. Straight approaches lacking such scientific approach are discouraged.

Indicative, but not complete, lists of topics covered in this focus session include:

  • Consensus in group decision-making
  • Consistency in fuzzy preference modeling
  • Missing preferences in fuzzy decision making
  • Aggregation of fuzzy preferences
  • Consensus measures
  • Consensus and fuzzy ontologies
  • Consensus software tools
  • Fuzzy decision in Web 2.0

FUZZ-IEEE-23 Bio-inspired Fuzzy Logic Approaches - Interdisciplinary Emergent Technologies (Cancelled)

Organized by Valentina E. Balas, Camelia Pintea, Ahmad Taher Azar, Mario Pavone, Rabie A. Ramadan and Nicolaie Popescu-Bodorin

In nature there are many examples that could help humanity to develop new projects, to improve and to solve some real life complex problems. The Bio-inspired Fuzzy Systems have the ability to include both natural computing and real life coefficients of uncertainty to keep in balance the solutions of the large-scale static and dynamic problems. The strategies of natural organisms (as ants, bees, nano-bots, swarms, flocks etc.) include adaptation and learning based on environmental changes, incomplete input information and the presence of noise. That is why Artificial Intelligence uses bio-inspired techniques, like ant colonies, artificial immune systems, swarm intelligence, neural networks, evolutionary computation, and not at last fuzzy logic to solve difficult problems.

Scope and Topics

The aim of this special session is provide an opportunity for international researchers to share and review recent advances in the foundations, integration architectures, and applications of Bio-inspired Fuzzy Logic systems in Pattern Recognition, Bioinformatics and computational biology, Healthcare, Industry, Microelectronics, Transportation, Green Logistics, Social Network, Web services, Cloud Computing and other domains. The topics of interest include, but are not limited to:

  • Fuzzy logic approaches in Evolutionary Computation, Swarm Intelligence, Genetic Algorithms, Particle Swarm Optimization, Ant Colony Optimization,Artificial Immune Systems and other natural computing systems
  • Uncertainty theory, programming, calculus and processes
  • Adaptive fuzzy pattern recognition
  • Learning based on Fuzzy Rule-Based Systems
  • Fuzzy-Neural and Hybrid schemes in adaptive estimation and control
  • Neuro-fuzzy technologies for medical and bioengineering applications
  • Bio-inspired fuzzy logic controllers for power system stabilizers
  • Agent based modeling and fuzzy logic
  • Multiobjective Bio-inspired Fuzzy Systems
  • Bio-inspired Fuzzy Clustering, Image Classification
  • Computational complexity
  • Microelectronics for Fuzzy and Bio-inspired Systems
  • Bio-inspired fuzzy models applied to cloud computing, transportation problems, systems automation, supply chain management, energy management systems, medicine, in robotics (bots / nano-bots), in social network and web services, complex data analysis: preprocessing and processing and other real life static and dynamic problems

FUZZ-IEEE-24 Medical Image Analysis based Computational Intelligence Techniques (Cancelled)

Organized by Nilanjan Dey, Amira S. Ashour, Dana Balas Timar and Valentina Emilia Balas,

Recently, the researchers’ intensive focus is attracted to medical image analysis studies. Neural network applications in computer-aided diagnosis (CAD) signify the foremost stream of computational intelligence in medical imaging. Moreover, neural networks are capable to optimize the inputs/ outputs relationship via distributed computing, training, and processing. This leads to reliable desired solutions by specifications, and medical diagnosis.

In the medical domain, the relation between accurate diagnosis and treatment can be assessed. Through medical imaging modalities physicians are able to collect/ measure information in the form of signals and/ or images that replicate the anatomical structure as well as the human body function. This field takes compensation of computer progress. In sake of effective diagnosis, computer aided systems become a must to construe and combine the acquired images for the purpose of diagnosis and intervention.

In spite of the success achieved by neural network in the medical domain, constructing multilayer neural networks includes challenging optimization problems.This session focuses on computational intelligence with neural networks covering medical image segmentation, registration, and edge detection for medical image analysis. In addition, computer-aided detection/ diagnosis with precise coverage on cancer screening, and other applications gives a global view on the variety of neural network applications and their potential for further research and developments.

Consequently, this special session is designed to allow the researchers, designers and developers to publish innovative and state-of-the art algorithms and architectures for medical image analysis based neural network with computational computing techniques.

Scope and Topics

The aim of this special session is to explore the existing neural network with computational intelligence to develop new efficient algorithms, to discussion of the existing problems/ challenges, and to propose solutions, and to collaborate for promising future research direction in the medical domain analysis based computational computing and machine learning. The topics of interest include:

  • Fuzzy logic and Neural networks in diagnosis in medicine
  • Fuzzy logic, Neural Networks/ Machine Learning in medicine- Theory and Application
  • Medical imaging modalities
  • Medical image and signal analysis
  • Visualization techniques
  • Segmentation and classification of medical images
  • Computer aided diagnosis (CAD) systems
  • Theoretical analysis of Neural Networks
  • Supervised learning and Unsupervised learning
  • Semi-supervised learning
  • Evolutionary Computation/ genetic algorithms in Neural Networks
  • Computational Intelligence in medicine
  • Classification using non-standard metrics
  • Machine learning methods in cancer
  • Computational Intelligence in Biomedicine
  • Recent developments in classification/clustering algorithms
  • Empirical analysis of Computational Intelligence algorithms for medical image analysis
  • Novel trends for medical image analysis based neural network with computational intelligence.

FUZZ-IEEE-25 Complex Fuzzy Sets and Logic

Organized by Scott Dick

Complex fuzzy sets are an extension to type-1 fuzzy sets in which membership grades are complex-valued. Likewise, complex fuzzy logic is an isomorphic family of multi-valued logics whose truth values are complex numbers. In the ten years since these concepts were first proposed, further theoretical investigations and a number of applications have made complex fuzzy sets and logic a lively and growing research area.

Scope and Topics

This special session will provide a forum to consolidate the community of researchers in this area, share our current ideas, reflect on future directions, and communicate our ideas and vision to the larger Computational Intelligence community. As such, we welcome submissions on all aspects of complex fuzzy sets or complex fuzzy logic, including but not limited to:

  • Theory of complex fuzzy logic
  • Complex fuzzy sets
  • Complex fuzzy inferential systems
  • Elicitation of complex fuzzy rules
  • Machine learning for complex fuzzy inferential systems
  • Hybridizations of complex fuzzy sets and logic with other CI technologies
  • Data mining with complex fuzzy sets and logic
  • Applications of complex fuzzy sets and logic
  • Complex fuzzy logic hardware

FUZZ-IEEE-26 From Type-1 to Type-n Fuzzy Systems Modeling

Organized by Mohammad H. Fazel Zarandi, Jerry Mendel and Burhan Turksen

In many real world problems, we encounter high uncertain information and knowledge based on which decision making should be considered. In such situations, type-1 or interval type-2 fuzzy set theory can be used to model and solve problems with vague information and knowledge. In some problems, the information is too vague to model the problem with either type-1 or intervalvalued type-2 fuzzy sets, so full type-2 or higher level fuzzy sets are used to model these systems. In full type-2 fuzzy sets, each element is represented by two memberships, which are named primary and secondary memberships. This fact shows the capability of full type-2 fuzzy sets in containing and representing more information than type-1 or interval fuzzy sets. Because of that capability, real problems with higher degree of uncertainty are solvable. Hence, there is an increasing need to do more research in the area of type-2 or higher level fuzzy systems and modeling to manage very uncertain problems.

Scope and Topics

Regarding to the increasing need for developing type-2 or higher level fuzzy systems, this session welcomes the researchers and papers in the area of theory and applications of type-2 and higher level fuzzy systems. The topics of this session include but are not limited to the following areas:

  • Operations on type-2 fuzzy sets
  • Interval and full type-2 fuzzy systems
  • Operations on type-n fuzzy sets
  • Type reduction methods
  • Direct and indirect approaches in type-2 fuzzy systems and modeling
  • Interval and full Type-2 fuzzy pattern recognition
  • Supervised and Unsupervised Type-2 fuzzy learning methods
  • Type-2 fuzzy functions
  • From Type-1 to Type-n fuzzy membership function generation techniques
  • Real-World applications of higher level (type-2 and more) fuzzy systems in control, robotics, finance, economics, engineering, medicine, image processing, and computer vision.

FUZZ-IEEE-27 Fuzzy Logic and Computational Intelligence Applications in Construction Engineering and Management (Cancelled)

Organized by Aminah Robinson Fayek and Chrysostomos Stylios

Construction engineering and management research has seen significant growth in fuzzy logic and computational intelligence applications to solve numerous problems. Fuzzy logic and computational intelligence have been used to model subjective information, handle uncertainty, and address the lack of comprehensive data sets available for modeling in construction engineering and management. In the construction domain, fuzzy logic has been combined with other soft computing techniques and computational intelligence methods to model, simulate, and create hybrid dynamic systems. This session will focus on recent advances and applications of fuzzy logic and computational intelligence techniques for applications related to planning and scheduling, estimating and bidding, productivity, organization competency, project control, structuring projects, process improvement, risk analysis, and others. In particular, challenges related to applying fuzzy logic in the construction engineering and management domain will be discussed and ideas generated on how to adapt fuzzy logic and fuzzy hybrid techniques to better suit construction applications.

Scope and Topics

The main topics of this special session include, but are not limited to, the application of the following approaches to construction engineering:

  • Fuzzy Expert Systems
  • Fuzzy Inference Model Prediction
  • Soft Computing for Risk Assessment
  • Classification in Construction
  • Modeling and Simulation of Construction Systems
  • Prediction Approaches in Construction Engineering
  • Decision Making in Construction Engineering
  • Neuro-fuzzy modeling
  • Interval Analysis for Construction Systems
  • Neuro-fuzzy Inference Systems
  • Fuzzy Cognitive Maps
  • Evolutionary Fuzzy Neural Inference Systems
  • Fuzzy Reasoning
  • Fuzzy Agent Based Modeling
  • Fuzzy System Dynamics
  • Reliability Analysis
  • Risk Analysis and Decision Making
  • Uncertainty Propagation
  • Performance of Construction Systems
  • Construction Productivity and Performance
  • Organizational Competency and Performance
  • Quality Management
  • Project Management and Operations Research

FUZZ-IEEE-28 Fuzzy-based Methods for Machine Learning: Data Preprocessing, Learning Models and Their Applications

Organized by Mikel Galar, Bartosz Krawczyk and Isaac Triguero

The aim of this special session is to serve as a forum for the exchange of ideas and discussions on recent and new trends regarding intersections between fuzzy systems and machine learning methods. Machine learning is a very active research field because of the huge number of real-world applications that can be addressed by this field of research. There are many contemporary problems, besides the canonical classification, regression or clustering, that require special focus and development of novel and efficient solutions. Such challenges include the problem of imbalanced data, learning on the basis of low quality and noisy examples, multi-label and multi-instance problems, or having limited access to object labels at the training phase, among others.

Learning methods based on Soft Computing techniques are widely used to face the aforementioned challenges with promising results. Fuzzy systems have demonstrated the ability to provide at the same time interpretable models understandable by human beings, as well as highly accurate results. Moreover, fuzzy-based techniques are of great interest when dealing with low quality or noisy data as they provide a framework to manage uncertainty. Evolutionary computation is a robust technique for optimization, learning and preprocessing tasks. They can adapt the model parameters for each problem to obtain a highly accurate system forming a good synergy with fuzzy approaches.

We encourage authors to submit original papers as well as preliminary and promising works in the topics of this special session.

Scope and Topics

The aim of the session is to provide a forum for the exchange of ideas and discussions on Soft Computing techniques and algorithms for machine learning, in order to deal with the current challenges in this topic. The special session is therefore open to high quality submissions from researchers working in learning problems using soft computing techniques. The topics of this special session include fuzzy models for handling data-level difficulties and improving machine learning methods in areas such as:

  • Supervised / Unsupervised / Semi-supervised learning
  • Feature Selection / Extraction / Construction
  • Instance Selection / Generation
  • Data streams and concept drift
  • Big data mining
  • Imbalanced learning
  • Multi-label \ Multi-instance learning
  • Feature and label noise
  • Kernels and Support Vector Machines
  • Ensemble learning
  • Evolutionary fuzzy systems
  • One-class classification / Learning from positive and unlabeled samples
  • Manifold Learning
  • Real-world applications e.g., in medical informatics, bioinformatics, social
  • networks, biometry, etc.

FUZZ-IEEE-29 Recent Advances in Fuzzy Control System Design and Analysis

Organized by Jun Yoneyama and Zsofia Lendek

The aim of this special session is to present the state-of-the-art results in the area of theory and applications of fuzzy control system design and analysis, and to get together well-known and potential researchers in this area. Fuzzy control system design and analysis provide a systematic and efficient approach to controlling of nonlinear plants and analysis of nonlinear control systems. Fuzzy control system has been employed to deal with a wide range of nonlinear control systems. A number of results on this area have appeared in the literature. However, there is still room for improvement of the existing results in order to propose new techniques for control of nonlinear systems. In the proposed special session, the focus is mainly on the fuzzy control system design and analysis with emphasis on the theory and applications. The important problems and difficulties on the fuzzy control systems will be addressed, their concepts will be provided and methodologies will be proposed to take care of the nonlinear systems using the fuzzy control system approaches.

Scope and Topics

The main topics of this special session include, but are not limited to:

  • Takagi-Sugeno fuzzy control system
  • Uncertain fuzzy system
  • Fuzzy hybrid system
  • Fuzzy switching system
  • Fuzzy time-delay system
  • Fuzzy stochastic system
  • Fuzzy polynomial system
  • Type-2 fuzzy control system
  • Stability analysis of Takagi-Sugeno fuzzy system
  • Nonlinear control design based on Takagi-Sugeno fuzzy system
  • Predictive control
  • Robust control
  • Sampled-data control
  • Filtering
  • Sliding mode control and observer

FUZZ-IEEE-30 Fuzzy Approaches for Advanced Manufacturing

Organized by Luka Eciolaza and George Panoutsos

Modern manufacturing environments are evolving considerably in order to adopt new ICT technologies and exploit their full potential to develop the so called factories of the future. Some of the main objectives consist on: (i) making sustainable manufacturing processes (highly efficient, productive, quality and accurate adaptive production processes), (ii) integrating human expert knowledge with the technology (iii) reducing the use of resources and generation of waste, (iv) opening new markets. The use of digital technologies throughout the manufacturing value chain plays a key role in order to achieve these goals.

Advanced Manufacturing implies an advanced degree of automation, autonomy and digitization within industrial processes and factories. Thus, advances in electronics and information technologies are considered key enabling technologies which are driving the transformation of current manufacturing systems towards the so called “Intelligent Factories”. The volume of data generated and archived in the manufacturing processes (design, simulation, monitoring, quality control, maintenance, etc.), represents a rich source of information which could potentially provide deep insight into the underlying physical processes and could also be used for process optimization. However, acquired process information is usually heterogeneous, complex, and with various degrees of uncertainty. Thus, intelligent data processing and analysis is an essential mechanism in order to extract useful knowledge models for their use in decision making.

Scope and Topics

The goal of this special session is to provide an insight into state of the art use of fuzzy logic based solutions in advanced manufacturing environments. These solutions should target mainly applications of: product design optimization, new manufacturing architectures for flexible manufacturing, product lifecycle management (PLM), zero­ defect manufacturing, additive manufacturing, maintenance services, computer aided monitoring and quality non­destructive testing (NDT), collaborative manufacturing environments. Potential topics of interest include but are not limited to:

  • Fuzzy Control
  • Data Mining, classification and information fusion
  • Incremental learning ­ Self­Learning systems
  • Modeling, Control, and Optimization
  • Decision Support Systems
  • Autonomous systems
  • Fault­tolerant control
  • Human­Centric Systems
  • Machine to Machine (M2M) communications
  • Predictive Process Control
  • Fault/Anomaly Detection and Clasification

FUZZ-IEEE-31 New Approaches to Fuzzy Web Intelligence

Organized by Giovanni Acampora, Chang-Shing Lee, Trevor Martin and Marek Reformat

Web intelligence is the area of scientific research and development that explores the roles and makes use of artificial intelligence and information technology methodologies for enabling the design and implementation of new products, services and frameworks that are empowered by the World Wide Web. In particular, Web intelligence achieves this goal through a combination of digital analytics, which examines how website visitors view and interact with a site’s pages and features, and business intelligence, which allows a corporation’s management to use data on customer purchasing patterns, demographics, and demand trends to make effective strategic decisions. As an example, search engines are one of the Internet applications that better benefit from this innovative method. Thanks to the aforementioned combination of technologies, Web Intelligence enables the implementation of enhanced systems aimed at improving users' experience in using and manipulating web resources, and companies' activities in deploying profiled and personalised contents and services. However, the imprecise and vague nature of World Wide Web, due to the large amount of information online and the different types of interaction that users and companies can have with this information, requires a new vision of web intelligence in which the treatment of uncertainty is a key factor: Fuzzy Web Intelligence. Indeed, recent literature review suggests that more and more successful developments in Web Intelligence are being integrated with fuzzy sets to enhance smart functionality such as web search systems by fuzzy matching, Internet shopping systems using fuzzy multi-agents, product recommender systems supported by fuzzy measure algorithms, e-logistics systems using fuzzy optimisation models; online customer segments using fuzzy data mining, fuzzy case-based reasoning in e-learning systems, and particularly online decision support systems supported by fuzzy set techniques. In light of the these observations, this special session is intended to form an international forum presenting innovative developments of fuzzy set applications in Web-based support systems. The ultimate objective is to bring well-focused high quality research results in Fuzzy Web Intelligence systems with intent to identify the most promising avenues, report the main results and promote the visibility and relevance of fuzzy sets.

Scope and Topics

The main topics of this special session include, but are not limited to:

  • Fuzziness in Web-based group support systems
  • Fuzziness in Web-based decision support systems
  • Fuzziness in Web-based personalised recommender systems
  • Fuzziness in Web-based knowledge management systems
  • Fuzziness in Web-based customer relationship management
  • Fuzziness in Web-based tutoring systems
  • Fuzzy Technology in e-Business intelligence
  • Fuzzy Technology for search engine design
  • Fuzzy Technology in e-Commerce intelligence
  • Fuzzy Technology in e-Government intelligence
  • Fuzzy Technology in e-Learning intelligence
  • Fuzzy Technology in e-Health intelligence
  • Fuzzy Ontologies
  • Fuzzy Sets and Semantic Web Applications
  • Web-based technologies for Fuzzy Reasoning
  • Fuzzy Markup Language and Applications

FUZZ-IEEE-32 Fuzzy Pattern Recognition for Big Data Modeling and Data Mining

Organized by Mohammad H. Fazel Zarandi, Oscar Castillo , Burhan Turksen and Behshad Lahijanian

In recent years, “Big Data” and “Data Mining” have become new ubiquitous terms. Big data and data mining are transforming science, engineering, medicine, healthcare, finance, business, and ultimately society itself. On the other hand, pattern recognition focuses on the recognition and regularities in data and tries to classify observations. Classification, data clustering, regression, sequence labeling, and parsing, etc. are some pattern recognition methods.

By consideration of pattern recognition techniques, data processing and making intelligent decisions on different area has been facilitated because of its capability of discovering patterns from data, there is an increasing need to do more research in the area of pattern recognition and data mining to handle complex problems. Regarding to the increasing need for developing pattern recognition techniques to manage the complexity of systems.This session welcomes researchers and papers in the different areas of theory and applications of pattern recognition, data mining, intelligent agents, etc.

Scope and Topics

The main topics of this special session include, but are not limited to:

  • Classification methods in pattern recognition (crisp, type-1 fuzzy, and type-2 fuzzy)
  • Clustering methods in pattern recognition (crisp, type-1 fuzzy, and type-2 fuzzy)
  • Graph-Based techniques (crisp, type-1 fuzzy, and type-2 fuzzy)
  • Sequential Data Analysis
  • Statistical, Structural and Syntactic Pattern Recognition
  • Kernel methods in pattern recognition
  • Sparse Kernel Machines
  • Learning methods for fuzzy pattern recognition (fuzzy neural networks, Support Vector Machines, Relevance Vector Machines, etc.)
  • Intelligent agent systems (crisp, type-1 fuzzy, and type-2 fuzzy)
  • Image perception and Observer Performance
  • Computer-Aided Diagnosis and Quantitative Image Analysis
  • Big Data pattern analysis
  • Hidden learning methods in pattern recognition and data mining
  • Fuzzy Data Mining methods and applications
  • Real-World applications of type-1 and type-2 fuzzy pattern recognition and data mining in intelligent agents, image processing, biological, computational biology, medicine, voice recognition, and computer vision,etc.

FUZZ-IEEE-33 Information Fusion and Fuzzy Linguistic Decision Making

Organized by Luis Martínez, Rosa M. Rodríguez and Francisco Herrera

Decision Making is an inherent mankind task related to intelligent and complex activities in which human beings face situations where they must choose among different alternatives by means of reasoning and mental processes. Such decision situations usually involve different types of uncertainty according to their nature. The fusion of information can reduce uncertainty and facilitate the decision making process because it associates, correlates and combines information from multiple sources to provide a relevant and timely view of the situation.

Therefore, information fusion in decision making has been widely studied from different points of view according to the framework in which it should be developed. However, there are still different open challenging problems related to information fusion and decision making because of the necessity of dealing with either novel decision making problems with new types of uncertainty and their modelling or with the advances in information fusion that imply improvements regarding previous approaches.

Additionally, many real decision situations are defined under uncertain contexts with imprecise information, in which it is straightforward the use of linguistic information. Fuzzy linguistic approach based models and Computing with Words (CW) provides the tools and methodology to deal with words. CW emulates human cognitive processes to improve decision solving processes under uncertainty. Consequently, information fusion processes, fuzzy linguistic approach and CW have been applied as modelling and computational basis for linguistic decision making, because it provides tools close to human beings reasoning processes related to decision making, which improve and facilitate the resolution of decision making under uncertainty as linguistic decision making.

All Information Fusion, Decision Making, Fuzzy Linguistic Approach and Computing with Words have recently attracted much attention in which, novel mathematical foundations and new decision models raised to be applied in different decision fields such as multi-criteria decision making, decision analysis, evaluation processes, consensus reaching processes, etc.

Scope and Topics

This invited session aims at providing an opportunity for researchers working in both research areas to discuss and to share their new ideas, original research results and practical experiences. More specifically, we expect you to have any contribution with the focus on the use of linguistic modelling in decision making. The topics of this special session are as follows:

  • Fusion Methods for Linguistic Decision making
  • Linguistic expression domains to represent preferences
  • Linguistic hesitant for modeling preference
  • Multi-criteria and group decision making
  • Selection and consensus models with linguistic information
  • Combining Heterogeneous Information in Decision Making
  • Multi-Level Fusion for Decision Making
  • Large Scale Decision Making
  • Fusing Linguistic Information in Decision Making
  • Intelligent Decision Support Systems
  • Dynamic Decision Making
  • Context-Based Information Fusion
  • Fusion in Networked Systems
  • Linguistic decision making in Engineering evaluation, resource management and transfer, Industry applications, sensory evaluation, evaluation and recommendation, Investments applications and risk assessment

FUZZ-IEEE-34 Methods of Operations Research For Decision Support Under Uncertainty

Organized by Jaroslav Ramik, Radomir Perzina, Elena Mielcova, Jiri Mazurek, Hana Tomaskova and Richard Cimler

Decision analysis based on uncertain data is natural in many real-world applications, and sometimes such an analysis is inevitable. In the past years, researchers have proposed many efficient operations research models and methods, which have been widely applied to real-life problems, such as finance, management, manufacturing, supply chain, transportation, among others. This special session aims to provide a forum for advancing the analysis, understanding, development, and practice of uncertainty theory and operations research for solving economic, engineering, management, and social problems

Scope and Topics

The goal of this special session is to provide an excellent forum for the discussion of the latest methods of operations research. The scope of the session will be focused on development of new methods of multiple criteria decision making under conditions of uncertainty and risk based on possibility theory and fuzzy sets theory. A new theory based on fuzzy relations and duality principle will be welcome. We invite the submission of high-quality, original and unpublished papers in this area. Interesting applications are welcome. The topics of interest include, but are not limited to:

  • Preference theories under uncertainty
  • Pair-wise comparison methods
  • Reciprocity and consistency of preference matrix
  • Fuzzy extensions of the well known methods (AHP etc.)
  • Intuitionistic, interval and fuzzy settings
  • Application of special SW for multi-criteria decision analysis
  • Other real-world applications

FUZZ-IEEE-35 Moving Towards Neutrosophic Logic

Organized by Swati Aggarwal,

Real life problems often calls for decision making under uncertainty—meaning we have to make a choice based on incomplete and indeterminate input data, often with unknown outcomes too. Researchers experimenting in the field of automated decision support systems make provisions of handling different types and potential sources of uncertainty whilst harmonizing the numerous aims of the system.

Real world problems have been effectively modeled using fuzzy logic that gives suitable representation of real-world data/information and enables reasoning that is approximate in nature. It is quite uncommon that the inputs captured by the fuzzy models are 100% complete and determinate. Though, humans can take intelligent decisions in such situations but fuzzy models require complete information. Incompleteness and indeterminacy in the data can arise from inherent non-linearity, time-varying nature of the process to be controlled, large unpredictable environmental disturbances, degrading sensors or other difficulties in obtaining precise and reliable measurements. Neutrosophic logic is an extended and general framework for measuring the truth, indeterminacy and falsehood-ness of the information. It is effective in representing different attributes of information like inaccuracy, incompleteness and ambiguous, thus giving fair estimate about the reliability of information.

This special session aims to provide a platform, where researchers coming from academia and industry can exhibit the varied practices of handling uncertainty in varied domains through the concepts of Neutrosophic Logic, communicate the connections amongst procedure and practice, and explain the contemporary case studies in different areas of application.

Scope and Topics

Topics of interest include, but are not limited to:

  • Neutrosophy
  • Neutrosophic Logic, Neutrosophic Sets
  • Neutrosophic Computation
  • Neutrosophic Probability
  • Neutrosophic Statistics
  • Neutrosophic reasoning systems
  • Neutrosophic applications

FUZZ-IEEE-36 Fuzzy Social Network Analysis (Cancelled)

Organized by Susan Bastani, Mohammad Hossein Fazel Zarandi, Jerry Mendel and Mansoureh Naderipour

Nowadays, social networks analysis is one of the main research subjects in computational intelligence, computer science, and sociology. Its importance is growing every day with expansion of social media, networks, and technological advancement.

Social networks, especially those one that have typical and non-commercial applications, are places in the virtual world that introduce their people briefly and provide possibility of communication between themselves and their adherents in the various interest areas. Obviously, virtual social networks will become more important and popular in the future. With social networks, persons are not alone to find their adherents in various cases.

Determining and predicting communications within a network is the main interest of social networks scientists and researchers. In the real world social networks, the communications are not usually defined crisply. In other words, the human communications usually encounters with imprecision and vagueness. Fuzzy theory, specially type-2 fuzzy logic, is very powerful approach to model social networks and analysis different ties (strong or weak) between nodes of the graphs. Regarding the increasing need for developing fuzzy topics in Social networks, this session welcomes the researchers and papers in the area of theory and applications of fuzzy theory (type-1 and type-2) in Social networks.

Scope and Topics

The topics of this session include but are not limited to the following areas:

  • Type-1 and type-2 fuzzy techniques for the decentralizing online social networks
  • Understanding and predicting human behavior for social communities with Fuzzy methods
  • Application of Fuzzy theory in social networks
  • Associating human-centered concepts with social networks using Fuzzy sets
  • Fuzzy node centrality in social network
  • Fuzzy group central potential in social network
  • Fuzzy data mining for fuzzy social network analysis
  • The closeness centrality analysis of fuzzy social network
  • Fuzzy classification methods for detecting communities in social networks
  • Fuzzy clustering methods for detecting communities in social networks
  • Fuzzy techniques for discovering communities from social networks: methodologies and applications
  • Dynamic social network communities using fuzzy methods
  • Positional analysis in fuzzy social networks
  • Role assignment in fuzzy social network
  • Fuzzy relations in social network analysis: optimization and consensus evaluation

FIND US ON

line line

IEEE CEC Special Sessions


IEEE CEC 2016 Special Sessions

CEC-01 Optimization Methods in Energy Internet System

Organized by Rui Wang, Sanaz Mostaghim, Tao Zhang and Shengxi Yang

Due to the rapid industrialization and the scarcity of conventional energy resources such as coal and natural gas, it has become increasingly urgent to find effective and efficient ways for energy use. The “Energy Internet System (EIS)” is a peer to peer energy exchange and sharing network which effectively integrates different energy sources together, including both conventional energy resources and renewable energy sources like solar and wind, and has become a promising solution.

However, there are various optimization issues existed in EISs. For example, the optimal structure design of the EIS, the optimal control and management of energy exchange, and the optimal scheduling of energy flow among different nodes. Moreover, hybrid renewable energy systems are often used in an EIS. The design of HRES is effectively a multi-objective optimization problem, that is, multiple objectives (such as the lifetime system cost, carbon emissions, and the system reliability) that are to be optimized. Therefore, the need for researchers from both optimization side and energy side to develop more effective and efficient methods to tackle issues arise in Energy Internet Systems has become apparent.

The main aim of this special session is to bring together both experts and new-comers from either academia or industry to discuss new and existing optimization issues in an EIS, in particular, to cross- fertilizate between academic research and industry applications, and to stimulate further engagement with the user community.

Scope and Topics

Full papers are invited on recent advances in the development of EISs, new horizons, i.e., using multi- criteria decision making methods, for EIS design and/or management. In addition, we are interested in various studies discussing optimization issues in EISs or related real-world applications. You are invited to submit papers that are unpublished original work for this special session. The topics include, but are not limited to:

  • The optimal structure design of the EIS
  • The optimal control and management of energy exchange
  • The optimal scheduling of energy flow among different nodes
  • Large decision variables based EIS optimal design or/and management
  • Multi-objective optimal design of HRES
  • HRES optimal design under dynamic environments
  • HRES optimal design under uncertain environments
  • Robust optimal design of HRES
  • Optimization methods for other energy related real-world problems

CEC-02 Intelligent Evaluation of Complex Algorithms (IEOCA)

Organized by Carsten Mueller, Markus Brenkner and Andre Hofmeister

A global acting logistic company uses metaheuristics to optimize the time-consuming and complex computation of transport paths. The company is worldwide connected and has different locations in Hamburg, St. Petersburg, New York, Hong Kong and Shanghai. Sophisticated employees continuously improve the computation of transport paths and evolve new algorithms.

A flexible solution called "Intelligent Evaluation Of Complex Algorithms" (IEOCA), which facilitate the collaboration and intelligent evaluation of existing and newly generated algorithm, is installed on a secure cloud platform in Hamburg. IEOCA provides a highly flexible, scalable and component based three-layer-architecture. These layers are protected with secure X.509 certificates and build the base of a trusted company network. Furthermore, the layers are linked through dynamically configurable service channels and ensure an extremely high-performance data exchange.

Every algorithm is encapsulated as modular component and attached to the cloud platform for evaluation purpose. These components are based on interfaces and easy to develop as well as to maintain. A team of experts evaluates the quality of the algorithms with fixed methods and expensive reports. In addition, IEOCA provides an automatic evaluation monitor. This complex monitor observes the performance of each algorithm i.e. runtime, memory usage or result and replace naturally worse components.

The IEOCA framework consumes information straightforward from the productive system via an extremely fast and lightweight communication channel. This mechanism allows the company to generate the maximum of economic benefit and improvements directly affect the daily business. This workshop shows in an impressive use case the implemented platform IEOCA and provides an interesting insight in the architecture as well as in the performance to the participants.

Scope and Topics

The proposed special session aims to bring together theories and applications of a dynamic component-based software architecture to the intelligent evaluation of complex algorithms. Topics of interest include, but are not limited to:

  • Component-based framework
  • Distributed Software Architecture and EventBus
  • Dynamic Service Composition
  • Evaluation of metaheuristics
  • Statistical analysis and hypothesis testing

CEC-03 Nature-Inspired Constrained Single- and Many-Objective Optimization

Organized by Helio J. C. Barbosa, Yong Wang and Efren Mezura-Montes

In their original versions, nature-inspired algorithms for optimization such as evolutionary algorithms (EAs) and swarm intelligence algorithms (SIAs) are designed to sample unconstrained search spaces. Therefore, a considerable amount of research has been dedicated to adapt them to deal with constrained search spaces. The objective of the session is to present the most recent advances in constrained optimization for single-, multi-, and many-objective optimization, using different nature-inspired techniques.

Scope and Topics

The session seeks to promote the discussion and presentation of novel works related with (but not limited to) the following issues:

  • Novel constraint-handling techniques for single-, multi-, and many-objective optimization
  • Novel constraint-handling techniques for constrained dynamic optimization
  • Novel/adapted search algorithms for constrained optimization
  • Memetic algorithms in constrained search spaces
  • Parameter setting (tuning and control) in constrained optimization
  • Mixed (discrete-continuous) constrained optimization
  • Theoretical analysis and complexity of algorithms in constrained optimization
  • Performance evaluation of algorithms in constrained optimization
  • Expensive Constrained Optimization
  • Design of difficult and scalable test functions
  • Applications

CEC-04 Evolutionary Computer Vision

Organized by Mengjie Zhang, Vic Ciesielski and Mario Koppen

Computer vision is a major unsolved problem in computer science and engineering. Over the last decade there has been increasing interest in using evolutionary computation approaches to solve vision problems. Computer vision provides a range of problems of varying difficulty for the development and testing of evolutionary algorithms. There have been a relatively large number of papers in evolutionary computer vision in recent CEC and GECCO conferences. It would be beneficial to researchers to have these papers in a special session. Also, a special session would encourage more researchers to continue to work in this field and consider CEC a place for presenting their work.

Scope and Topics

The proposed special session aims to bring together theories and applications of evolutionary computation to computer vision and image processing problems. Topics of interest include, but are not limited to:

New theories and methods in different EC paradigms for computer vision and image processing including

  • Evolutionary algorithms such as genetic algorithms, genetic programming, evolutionary strategies and evolutionary programming;
  • Swarm Intelligence methods such as particle swarm optimisation, ant colony optimisation, and differential evolution;
  • Other approaches such as learning classifier systems, harmony search, and artificial immune systems. Cross-fertilization of evolutionary computation with other techniques such as neural networks and fuzzy systems is also encouraged.
Applications in computer vision and image processing including
  • Detection in noisy images
  • Image segmentation in biological images
  • Automatic feature extraction, construction and selection in complex images
  • Object identification and scene analysis for medical applications
  • Object detection and classification in security scenarios
  • Handwritten digit recognition and detection
  • Vehicle plate detection
  • Face detection and recognition
  • Texture image analysis
  • Automatic target recognition in military services
  • Gesture identification and recognition
  • Robot vision

CEC-05 Evolutionary Scheduling and Combinatorial Optimization

Organized by Su Nguyen, Yi Mei and Mengjie Zhang

Evolutionary Scheduling and Combinatorial Optimization is an active research area in both Artificial Intelligence and Operations Research due to its applicability and interesting computational aspects. Evolutionary techniques are suitable for these problems since they are highly flexible in terms of handling constraints, dynamic changes and multiple conflicting objectives.

Scope and Topics

This special session focuses on both theoretical and practical aspects of Evolutionary Scheduling and Combinatorial Optimization. Examples of evolutionary methods include genetic algorithm, genetic programming, evolutionary strategies, ant colony optimisation, particle swarm optimisation, evolutionary based hyper-heuristics, memetic algorithms.

  • Production scheduling
  • Timetabling
  • Vehicle routing
  • Transport scheduling
  • Grid/cloud scheduling
  • Project scheduling
  • 2D/3D strip packing
  • Space allocation
  • Multi-objective scheduling
  • Multiple interdependent decisions
  • Automated heuristic design
  • New real-world and innovative applications

CEC-06 Evolutionary Feature Selection and Construction

Organized by Bing Xue, Mengjie Zhang and Yaochu Jin

Many data mining and machine learning problems involve a large number of features/attributes, which leads to “the curse of dimensionality”. However, not all the features are essential since many of them are redundant or even irrelevant, and the “useful” features are typically not equally important. This problem can be solved by feature selection to select a small subset of original (relevant) features or feature construction to create a smaller set of high-level features using the original low-level features and mathematical or logical operators. Feature selection and construction are challenging tasks due to the large search space and feature interaction problems. Recently, there has been increasing interest in using evolutionary computation techniques to solve feature selection and construction tasks.

Scope and Topics

The theme of this special session is the use of evolutionary computation for feature reduction, covering ALL different evolutionary computation paradigms. The aim is to investigate both the new theories and methods in different evolutionary computation paradigms to feature reduction, and the applications of evolutionary computation for feature reduction. Authors are invited to submit their original and unpublished work to this special session.
Topics of interest include but are not limited to: 

  • Feature ranking/weighting
  • Feature subset ranking
  • Feature subset selection
  • Filter, wrapper, and embedded methods for feature selection
  • Multi-objective feature selection
  • Feature construction/extraction
  • Single feature or multiple features construction
  • Filter, wrapper, and embedded methods for feature construction
  • Multi-objective feature construction
  • Analysis on evolutionary feature selection and construction algorithms
  • Feature selection and construction in classification, clustering, regression, and other tasks
  • Feature selection and construction on high-dimensional and large-scale data
  • Hybridisation of evolutionary computation and neural networks, and fuzzy systems for feature selection and construction
  • Hybridisation of evolutionary computation and machine learning, information theory, statistics, mathematical modelling, etc., for feature selection and construction
  • Real-world applications of evolutionary feature selection and construction, e.g. images and video sequences/analysis, face recognition, gene analysis, biomarker detection, medical data classification, diagnosis, and analysis, hand written digit recognition, text mining, instrument recognition, power system, financial and business data analysis, et al.

CEC-07 Evolutionary Bilevel Optimization

Organized by Ankur Sinha and Kalyanmoy Deb

Bilevel optimization problems are special kind of optimization problems that involve two levels of optimization, namely upper level and lower level. The hierarchical structure of the problem requires that every feasible solution to the upper level problem should satisfy the optimality conditions of the lower level problem. Such a requirement makes bilevel optimization problems difficult to solve. These problems are commonly found in many practical problem solving tasks, which include optimal control, process optimization, game-playing strategy development, transportation problems, coordination of multi-divisional firms, machine learning and others. Due to the computation expense and other difficulties involved in handling such problems, they are often handled using approximate solution procedures. There is a need for theoretical as well as methodological advancements to handle such problems efficiently.

Scope and Topics

The special session on Bilevel Optimization will focus on the following topics:

  • Evolutionary algorithms for bilevel optimization problems
  • Evolutionary algorithms for multi-objective bilevel optimization problems
  • Approximate procedures to handle bilevel optimization problems
  • Hybrid approaches to handle bilevel optimization problems
  • Theoretical results on bilevel optimization problems
  • Bilevel Application Problems
  • Hierarchical decision making

CEC-08 Many-Objective Optimization

Organized by Hiroyuki Sato and Antonio Lopez Jaimes

Evolutionary algorithms are particularly suited to solve multi-objective optimization problems since they can obtain a set of non-dominated solutions to approximate Pareto front in a single run of the algorithm. So far, multi-objective EAs have been successfully applied mostly in two and three objectives problems. However, multi-objective EAs face several difficulties when we try to solve many-objective optimization problems, which optimize four or more objective functions simultaneously. At least, the following difficulties have been recognized in recent researches of evolutionary many-objective optimization.
(1) The convergence deterioration of solutions toward Pareto front
(2) The approximation of high dimensional entire Pareto front with a limited number of solutions in the population
(3) The presentation of obtained solutions in the high dimensional objective space and the decision making of a single final solution from them
(4) The search performance evaluation of search algorithms

Scope and Topics

This special session will focus on evolutionary many-objective optimization to tackle problems in many-objective optimization including the above mentioned difficulties and the below few topic (but not limited to):.

  • Algorithm design
  • Preference based search
  • Dimensionality reduction of the objective space
  • Benchmark problems
  • Visualization of high dimensional space and decision making
  • Search performance metrics

CEC-09 Evolutionary Computation for Nonlinear Equation Systems

Organized by Yong Wang, Zixing Cai, Qingfu Zhang and Crina Grosan

Nonlinear equation systems (NESs) frequently arise in many physical, electronic, and mechanical processes. Very often, a NES may contain multiple optimal solutions. Since all these optimal solutions are important for a given NES in the real-world applications, it is desirable to simultaneously locate them in a single run, such that the decision maker can select one final solution which matches at most his/her preference.

For solving NESs, several classical methods, such as Newton-type methods, have been proposed. However, these methods have some disadvantages in the sense that they are heavily dependent on the starting point of the iterative process, can easily get trapped in a local optimal solution, and require derivative information. Moreover, these methods aim at locating just one optimal solution rather than multiple optimal solutions when solving NESs. During the past decade, evolutionary algorithms (EAs) have been widely applied to solve NESs due to the fact that EAs are insensitive to the shapes of the objective function and easy to implement.

Solving NESs by EAs is a very important area in the community of evolutionary computation, which is challenging and of practical interest. However, systematic work in this area is still very limited. This is the first special issue in the IEEE CEC to facilitate the development of EAs for NESs.

Scope and Topics

The topics of this special issue include (but are not limited to):

  • Design of benchmark NES test instances
  • Theoretical and experimental analysis of the optimal solutions in NESs
  • Novel search engines in evolutionary computation for NESs
  • Novel techniques for transforming a NES into a kind of optimization problems
  • Locating multiple optimal solutions of NESs by multiobjective optimization methods
  • Real-world applications

CEC-10 Computational Intelligence in Aerospace Science and Engineering

Organized by Massimiliano Vasile, Chit Hong Yam, Victor Becerra and Edmondo Minisci

In an expanding world with limited resources and increasing complexity, optimisation and computational intelligence become a necessity. Optimisation can turn a problem into a solution and computational intelligence can offer new solutions to effectively make complexity manageable.

All this is particularly true in space and aerospace where complex systems need to operate optimally often in harsh and inhospitable environment with high level of reliability. In Space and Aerospace Sciences, many applications require the solution of global single and/or multi-objective optimization problems, including mixed variables, multi-modal and non-differentiable quantities. From global trajectory optimization to multidisciplinary aircraft and spacecraft design, from planning and scheduling for autonomous vehicles to the synthesis of robust controllers for airplanes or satellites, computational intelligence (CI) techniques have become an important – and in many cases inevitable – tool for tackling these kinds of problems, providing useful and non- intuitive solutions. Not only have Aerospace Sciences paved the way for the ubiquitous application of computational intelligence, but moreover, they have also led to the development of new approaches and methods.

In the last two decades, evolutionary computing, fuzzy logic, bio-inspired computing, artificial neural networks, swarm intelligence and other computational intelligence techniques have been used to find optimal trajectories, design optimal constellations or formations, evolve hardware, design robust and optimal aerospace systems (e.g. reusable launch vehicles, re-entry vehicles, etc.), evolve scheduled plans for unmanned aerial vehicles, improve aerodynamic design (e.g. airfoil and vehicle shape), optimize structures, improve the control of aerospace vehicles, regulate air traffic, etc.

Scope and Topics

This special session intends to collect many, diverse efforts made in the application of computational intelligence techniques, or related methods, to aerospace problems. The session seeks to bring together researchers from around the globe for a stimulating discussion on recent advances in evolutionary methods for the solution of space and aerospace problems.
In particular evolutionary methods specifically devised, adapted or tailored to address problems in space and aerospace applications or evolutionary methods that were demonstrated to be particularly effective at solving aerospace related problems are welcome.

  • Global trajectory optimization
  • Multidisciplinary design for space missions
  • Formation and constellation design and control
  • Optimal control of spacecraft and rovers
  • Planning and scheduling for autonomous systems in space
  • Multiobjective optimization for space applications
  • Resource allocation and programmatics
  • Evolutionary computation for Concurrent Engineering
  • Distributed global optimization
  • Mission planning and control
  • Robust Mission Design under Uncertainties
  • Intelligent search and optimization methods in aerospace applications
  • Image analysis for Guidance Navigation and Control
  • Autonomous exploration of interplanetary and planetary environments
  • Implications of emerging AI fields such as Artificial Life or Swarm Intelligence future space research
  • Intelligent algorithms for fault identification, diagnosis and repair
  • Multi-agent systems approach and bio-inspired solutions for system design and control
  • Advances in machine learning for space applications
  • Intelligent interfaces for human-machine interaction
  • Knowledge Discovery, Data Mining and presentation of large data sets

CEC-11 Automated Design: Hyper-heuristics and Metaheuristics

Organized by Nelishia Pillay and Rong Qu

Designing metaheuristics to solve problems can be time consuming, requiring many man hours. This involves making a number of design decisions such as parameter tuning, identifying moves or operators to use, deciding on the control flow of the algorithm or determining which low-level construction heuristics to use in the case of combinatorial optimization problems. In some cases it may be necessary to create new operators or algorithms or hybridize different metaheuristics to solve a problem. Hyper-heuristics and adaptive metaheuristics have proven to be effective for making some of these design decisions, thereby facilitating automated design. Hyper-heuristics have been successfully used for the selection and generation of low-level heuristics in solving various combinatorial optimization problems including timetabling, vehicle routing, packing problems amongst others and have also been applied to dynamic environments and multiobjective optimization. More recent trends in hyper-heuristic research have focused on the design of metaheuristics. Selection hyper-heuristics have been used for determining parameter values, choice of operators and control flow in metaheuristics, e.g. evolutionary algorithms and ant colonization, as well as for the hybridization of techniques, e.g. multiobjective evolutionary algorithms, different metaheuristics. Generation hyper-heuristics have been employed to create new operators for metaheuristics, e.g. selection and mutation operators. An emerging area in hyper-heuristics is hyper-hyper-heuristics, i.e. using hyper-heuristics to generate or design hyper-heuristics.

Scope and Topics

The aim of this special session is for researchers to present recent developments in the field thereby paving the way for future advancement. The main topics include but are not limited to:

  • Applications of selection and generative hyper-heuristics
  • Hyper-heuristics for metaheuristic design, e.g. parameter tuning, control flow, operator selection
  • Hyper-heuristics for the creation of new operators and algorithms
  • Hyper-heuristics for the derivation of hybrid methods, e.g. hybridization of metaheuristics
  • Hyper-heuristics for the design hyper-heuristics
  • Cross domain applications of hyper-heuristics
  • Parallelization of hyper-heuristics
  • Theoretical aspects of hyper-heuristics

CEC-12 Brain Storm Optimization Algorithms

Organized by Shi Cheng, Quande Qin, Yuhui Shi and Simone Ludwig

Swarm intelligence algorithm should have two kinds of ability: capability learning and capacity developing. The capacity developing focuses on moving the algorithm’s search to the area(s) where higher search potential may be obtained, while the capability learning focuses on its actually search from the current solution for single point based optimization algorithms and from the current population for population-based swarm intelligence algorithms. The swarm intelligence algorithms with both capability learning and capacity developing can be called as developmental swarm intelligence algorithms.

The capacity developing is a top-level learning or macro-level learning methodology. The capacity developing describes the learning ability of an algorithm to adaptively change its parameters, structures, and/or its learning potential according to the search states of the problem to be solved. In other words, the capacity developing is the search strength possessed by an algorithm. The capability learning is a bottom-level learning or micro-level learning. The capability learning describes the ability for an algorithm to find better solution(s) from current solution(s) with the learning capacity it possesses.

The Brain Storm Optimization (BSO) algorithm is a new kind of swarm intelligence, which is based on the collective behaviour of human being, that is, the brainstorming process. It is natural to expect that an optimization algorithm based on human collective behaviour could be a better optimization algorithm than existing swarm intelligence algorithms which are based on collective behaviour of simple insects, because human beings are social animals and are the most intelligent animals in the world. The designed optimization algorithm will naturally have the capability of both convergence and divergence.

The BSO algorithm is a good example of developmental swarm intelligence algorithm. A “good enough” optimum could be obtained through solution divergence and convergence in the search space. In the BSO algorithm, the solutions are clustered into several categories, and the new solutions are generated by the mutation of cluster or existing solutions. The capacity developing, i.e., the adaptation during the search, is another common feature of the BSO algorithms.

The BSO algorithm can be seen as a combination of swarm intelligence and data mining techniques. Every individual in the brain storm optimization algorithm is not a solution to the problem to be optimized, but also a data point to reveal the landscapes of the problem. The swarm intelligence and data mining techniques can be combined to produce benefits above and beyond what either method could achieve alone.

Scope and Topics

This special session aims at presenting the latest developments of BSO algorithm, as well as exchanging new ideas and discussing the future directions of developmental swarm intelligence. Original contributions that provide novel theories, frameworks, and applications to algorithms are very welcome for this Special Session. Potential topics include, but are not limited to:

  • Analysis and control of BSO parameters
  • Parallelized and distributed realizations of BSO algorithms
  • BSO for Multi-objective optimization
  • BSO for Constrained optimization
  • BSO for Discrete optimization
  • BSO algorithm with data mining techniques
  • BSO in uncertain environments
  • Theoretical aspects of BSO algorithm
  • BSO for Real-world applications

CEC-13 Evolutionary Computation and Big Data

Organized by Shi Cheng, Yuhui Shi, Yaochu Jin and Bin Li

Nowadays, big data has been attracting increasing attention from academia, industry and government. Big data is defined as the dataset whose size is beyond the processing ability of typical databases or computers. Big data analytics is to automatically extract knowledge from large amounts of data. It can be seen as mining or processing of massive data, and “useful” information can be retrieved from large dataset. Big data analytics can be characterized by several properties, such as large volume, variety of different sources, and fast increasing speed (velocity). It is of great interest to investigate the role of evolutionary computing (EC) techniques, including evolutionary algorithms and swarm intelligence algorithms for the optimization and learning involving big data, in particular, the ability of EC techniques to solve large scale, dynamic, and sometimes multi-objective big data analytics problems.

Scope and Topics

This special session aims at presenting the latest developments of EC techniques for big data problems, as well as exchanging new ideas and discussing the future directions of EC for big data. Original contributions that provide novel theories, frameworks, and solutions to challenging problems of big data analytics are very welcome for this Special Session. Potential topics include, but are not limited to:

  • High-dimensional and many-objective evolutionary optimization
  • Big data driven optimization of complex engineering systems
  • Integrative analytics of diverse, structured and unstructured data
  • Extracting new understanding from real-time, distributed, diverse and large-scale data resources
  • Big data visualization and visual data analytics
  • Scalable, incremental learning and understanding of big data
  • Scalable learning techniques for big data
  • Big data driven optimization of complex systems
  • Human-computer interaction and collaboration in big data
  • Big data and cloud computing
  • Cross-connections of big data analysis and hardware
  • Big data techniques for business intelligence, finance, healthcare, bioinformatics, intelligent transportation, smart city, smart sensor networks, cyber security and other critical application areas
  • MapReduce implementations combined with evolutionary computation or swarm intelligence approaches

CEC-14 Fireworks Algorithm and Its Application on Big-Data

Organized by Ying Tan and Liangjun Ke

Big data contains huge amount of data and information and is worth researching in depth. Big data, also known as massive data or mass data, referring to the amount of data involved that are too great to be interpreted by a human. The Obama administration invested nearly two hundred million US dollars on the program of "Big Data Research and Development Initiative", aiming to protect the national security. In addition, sociologists use the big data from social interaction network to analyze the human behavior, communication methods.

However, the methods to process big data are ineffective. Currently, the suitable technologies include A/B testing, crowdsourcing, data fusion and integration, genetic algorithms, machine learning, natural language processing, signal processing, simulation, time series analysis and visualization. But real or near-real time information delivery is one of the defining characteristics of big data analytics. It is important to find new methods to enhance the effectiveness of big data.

Fireworks algorithm (FWA) achieved a great success on solving many complex optimization problems effectively. FWA has a unique search manner in the solution space and is a strong capability to solve optimization problems. It has many effective variants and huge amount of successful applications. Moreover, FWA is suitable for parallelization and works significantly better than other SI algorithms, such as particle swarm optimization, ant colony optimization and genetic algorithm.

The main aim of this special session is to bring together both experts and new-comers from either academia or industry to discuss fireworks algorithm and its application, especially on the big-data application. However, both the improvements and the applications of FWA are welcome and acceptable for this special session.

Scope and Topics

Full papers are invited on recent advances in the development of FWA, i.e., FWA improvements and applications. In addition, we are interested in various studies on discussing processing big data issues by FWA. The session seeks to promote the discussion and presentation of novel works related with (but not limited to) the following issues:

  • Theoretical analysis of FWA
  • Algorithmic improvement of FWA
  • FWA for single-, multi-, and many-objective optimization
  • FWA for data mining and machine learning
  • FWA for big data and data analysis
  • Parallelized and distributed realizations of FWA
  • Varieties of Applications of FWA
  • FWA for Big-data Applications
  • Other applications

CEC-15 Evolutionary Optimization for Non-Convex Machine Learning

Organized by Yang Yu, Ke Tang and Jose A. Lozano

Sophisticated optimization problems lay in many machine learning tasks. These problems were commonly smartly relaxed as convex optimization problems. Although the relaxation allows an efficient optimization using mathematical programming methods, it often shifts the learning problem and loses some important properties (e.g., convex loss functions may sensitive to data noise). Evolutionary optimization provides a set of direct search tools that make it possible to solve non-convex optimization problems for machine learning. This special session intends to bring together researchers to report their latest progress and exchange experience in solving machine learning tasks better with evolutionary optimization methods.

Scope and Topics

The interest of this special session is on solving non-convex optimization problems in machine learning with the methodologies related to evolutionary optimization, such as evolutionary algorithms, swarm intelligence algorithms, cross-entropy methods, Bayesian optimization. The topics cover a broad range of machine learning tasks including (but not limited to):

  • Supervised, semi-supervised, and multi-label learning
  • Learning deep models
  • Representation learning, sparse learning, dimension extraction
  • Reinforcement learning
  • Multi-instance learning
  • Cost-sensitive and imbalanced learning
  • Unsupervised learning and clustering
  • Parameter tuning

CEC-16 Fitness Landscape Analysis in Practice

Organized by Katherine M. Malan and Andries P. Engelbrecht

Since the notion of a fitness landscape was introduced by Sewell Wright in 1932, fitness landscapes have been studied by evolutionary biologists to better understand how evolution occurs in nature. In a similar way, researchers in evolutionary computation (EC) have used fitness landscapes to better understand the evolutionary process of search. Studies have ranged from theoretical models of fully enumerated combinatorial landscapes to the prediction of algorithm performance based on approximate fitness land- scape characteristics. Fitness landscape analysis is a growing field in the EC community, but research has been scattered in widely different publications and conferences. Research papers in fitness land- scapes are often incorporated into theoretical tracks of conferences, even when the research is focussed on the practical application of fitness landscape analysis. Alternatively, papers on fitness landscapes may appear alone in specific algorithm tracks, such as swarm intelligence or genetic programming.

The aim of this special session on fitness landscapes is to provide an opportunity to not only bring fitness landscape analysis researchers together at CEC 2016, but also to publish the most recent work in a dedicated track in the proceedings. In addition, the special session should be of interest to re- searchers and practitioners interested in practically applying fitness landscape analysis techniques to better understand problems and algorithm behaviour.

Scope and Topics

For this special session on fitness landscapes we invite researchers to submit unpublished work specifically focussing on the practice of fitness landscape analysis. Topics of interest include, but are not limited to:

  • Analysis of algorithm performance in relation to fitness landscape characteristics.
  • Practical techniques for characterising the features of combinatorial problems with large search spaces or approximating the features of continuous search spaces.
  • Practical measures for characterising dynamic and co-dynamic landscapes.
  • Practical analysis of the fitness landscapes of constrained optimisation problems.
  • Online fitness landscape analysis for the characterisation of problems during search.
  • Analysis of benchmark problem suites using fitness landscape techniques.
  • Generation of new benchmark problems with particular fitness landscape characteristics.
  • Analysis of the fitness landscapes of specific classes of problems or real-world optimisation problem instances to provide insight into algorithm behaviour or to highlight challenges in the practical application of fitness landscape analysis.

CEC-17 Computationally Expensive Optimization

Organized by Karthik Sindhya, Handing Wang, Markus Olhofer and Yaochu Jin

High fidelity CAE models result in very detailed and precise information on the system at hand and offer a huge potential for the utilization of numerical optimization methods like evolutionary computation. However a big challenge is the resulting high computational cost of the models which often even hinders the application of evolutionary optimization and design methods. In the literature various methods are proposed to tackle the problem like for example approximation methods or surrogates which are used to reduce computationally expensive optimization problems. However this involves many challenges in tailoring the approximation/surrogate to each problem in question. The aim of this session is to bring together researchers from evolutionary algorithms who deal with computationally expensive problems. This is an ideal platform for researchers and practitioners to interact and present ideas for handling computationally expensive problems.

Scope and Topics

The topics include (but not limited to):

  • Reducing of function evaluations using function, fitness, problem approximation and multi-level approaches
  • Surrogate based methods in optimization
  • Decision maker’s preference learning and many objective optimization including
  • objective and parameter reduction and interactive methods
  • Parallelization/grid/cloud of evolutionary algorithms
  • Data-driven optimization using big data and data analytics
  • Real world applications including multidisciplinary optimization

CEC-18 Big Optimization (BigOpt2016)

Organized by Hussein A. Abbass and Kay Chen Tan

Big Optimization (BigOpt) is the term we coin to differentiate optimization problems that rely on big data from classical large scale optimization. BigOpt problems involve thousands of variables and are normally expected to hide trends.

This special session is organized in conjunction with the BigOpt competition. How- ever, authors who do not contribute to the competition but have papers related to the optimization of big data problems are also encouraged to submit to the special session.

Scope and Topics

In this special session, several aspects of Evolutionary Algorithm and Nature Inspired Algorithms design can be considered, but not limited to the following:

  • Evolutionary and other Nature Inspired Algorithms Design for Big Optimization
  • Dimensionality reduction of big search spaces
  • Novel operator design for optimization problems with thousands or more variables
  • Single Objective Big Optimization
  • Multi Objective Big Optimization
  • Parallel Evolutionary and other Nature Inspired Algorithms for Big Optimization

CEC-19 Efficient Non-dominated Sorting and Pareto Approaches to Many-Objective Optimization

Organized by Xingyi Zhang, Ran Cheng, and Yaochu Jin

Solving many-objective optimization problems (MaOPs) has drawn increasing attention in the research community due to the fact that MaOPs cannot be solved efficiently using traditional MOEAs developed for solving multi-objective optimization problems with two or three objectives. Among various ideas adopted for many-objective optimization, one big challenge is to develop highly efficient and effective non-dominated sorting methods and Pareto-based multi-objective evolutionary algorithms (MOEAs) for MaOPs. The aim of this special session is to bring together researchers in the evolutionary computation community dedicated to solving MaOPs using non-dominated sorting and Pareto-based approaches to MaOPs.

Scope and Topics

Topics of the special session include, but are not limited to

  • Reducing the computational cost of non-dominated sorting for MaOPs
  • New Pareto dominance relationships for MOEAs to solve MaOPs
  • New convergence/diversity-related metrics combined with dominance comparison for MaOPs
  • Efficient Pareto-based MOEAs for large scale multi-objective optimization problems
  • Real world applications of efficient non-dominated sorting and/or Pareto-based MOEAs for MaOPs

CEC-20 When Evolutionary Computation Meets Data Mining

Organized by Zhun Fan, Xinye Cai, Chuan-Kang Ting and Qingfu Zhang

Many of the tasks carried out in data mining and machine learning, such as feature subset selection, associate rule mining, model building, etc., can be transformed as optimization problems. Thus it is very natural that Evolutionary Computation (EC), has been widely applied to these tasks in the fields of data mining (DM) and machine learning (ML), as an optimization technique.

On the other hand, EC is a class of population-based iterative algorithms, which generate abundant data about the search space, problem feature and population information during the optimization process. Therefore, the data mining and machine learning techniques can also be used to analyze these data for improving the performance of EC. A plethora of successful applications have been reported, including the creation of new optimization paradigm such as Estimation of Distribution Algorithm, the adaptation of parameters or operators in an algorithm, mining the external archive for promising search regions, etc.

However, there remain many open issues and opportunities that are continually emerging as intriguing challenges for bridging the gaps between EC and DM. The aim of this special session is to serve as a forum for scientists in this field to exchange the latest advantages in theories, technologies, and practice.

Scope and Topics

We invite researchers to submit their original and unpublished work related to, but not limited to, the following topics:

  • EC Enhanced by Data Mining and Machine Learning Concepts and/or Method
  • Data Mining and Machine Learning Based on EC Techniques
  • Data Mining and Machine Learning Enhanced Multi-Objective Optimization
  • Data Mining and Machine Learning Enhanced Constrained Optimization
  • Data Mining and Machine Learning Enhanced Memetic Computation
  • Multi-Objective Optimization and Rule Mining Problems
  • Knowledge Discovery in Data Mining via Evolutionary Algorithm
  • Genetic Programming in Data Mining
  • Multi-Agent Data Mining using Evolutionary Computation
  • Medical Data Mining with Evolutionary Computation
  • Evolutionary Computation in Intelligent Network Management
  • Evolutionary Clustering in Noisy Data Sets
  • Big Data Projects with Evolutionary Computation
  • Real World Applications

CEC-21 Evolutionary Computation for the Design of Digital Filters

Organized by Ranjit Singh Chauhan

During the last few decades, there has been an enhancement of using Evolutionary Computation (EC). Evolutionary Computation is a machine learning optimization and classification algorithm which is roughly based on the mechanism of evolutions such as Biological genetics and Natural selections. This session is to cover practical, theoretical and applied aspects of the engineering design of signals, with emphasis on Digital signals using Evolutionary algorithms. The term “signal” includes audio, video, speech, image, communication, geophysical, bio-medical, musical, and other signals. This session will discuss the issues of filtering, coding, transmitting, estimating, detecting, analyzing, recognizing, synthesizing, recording, and reproducing signals by analog or digital devices. Design, synthesis, integration, evaluation, standardization, and the development of algorithms evaluated towards building digital filters will be the primary focus.

Scope and Topics

The main aim of this session is to discuss the trends of evolutionary computation for the design of digital filters.
Topic/Area covered may be followings:

  • Signal analysis and synthesis
  • Image signal, speech signal, video signal, etc.
  • Application specific array processors
  • Design of digital filters and synthesizer
  • Generic algorithm for digital filters
  • Ant colony optimization for digital filters
  • Particle swarm optimization for digital filters
  • and other technologies/techniques

CEC-22 Evolutionary Robotics

Organized by Patricia A. Vargas, Dario Floreano, Joshua Auerbach, Micael Couceiro and Phil Husbands

Evolutionary Robotics (ER) aims to apply evolutionary computation techniques to automatically design the control and/or hardware of both real and simulated autonomous robots. Its origins date back to the beginning of the nineties and since then it has been attracting the interest of many research centres all over the world.

ER techniques are mostly inspired by existing biological architectures and Darwin’s principle of selective reproduction ofthe fittest. Evolution has revealed that living creatures are able to accomplish complex tasks required for their survival, thus embodying cooperative, competitive and adaptive behaviours.

Having an intrinsic interdisciplinary character, ER has been employed toward the development of many fields of research, among which we can highlight neuroscience, cognitive science, evolutionary biology and robotics. Hence, the objective of this special session is to assemble a set of high-quality original contributions that reflect and advance the state-of-the-art in the area of Evolutionary Robotics, with an emphasis on the cross-fertilization between ER and the aforementioned research areas, ranging from theroretical analysis to real-life applications.

Scope and Topics

Topics of interest include (but are not restricted to):

  • Evolution of robots which display minimal cognitive behaviour, learning, memory, spatial cognition , adaption or homeostasis
  • Evolution of neural controllers for robots, aimed at giving an insight to neuroscientists, evolutionary biologists or advancing control structures
  • Evolution of communication, cooperation and competition, using robots as a research platform
  • Co-evolution and the evolution of collective behaviour
  • Evolution of morphology in close interaction with the environment, giving rise to self-configurable , self-designing, self-healing, self-reproducing, humanoid and walking robots
  • Evolution of robot systems aimed at real-world applications as in aerial robotics, space exploration, industry, search and rescue, robot companions, entertainment and games
  • Evolution of controllers on board real robots or the real-time evolution of robot hardware
  • Novel or improved algorithms for the evolution of robot systems
  • The use of evolution for the artistic exploration of robot design

CEC-23 Memetic Computing

Organized by Liang Feng, Ferrante Neri and Yew-Soon Ong

Memetic Computing (MC) represents a broad generic framework using the notion of meme(s) as units of information encoded in computational representations for the purpose of problem-solving. In the literature, MC has been successfully manifested as memetic algorithm, where meme has been typically perceived as individual learning procedures, adaptive improvement procedures or local search operators that enhance the capability of population based search algorithms. More recently, novel manifestations of meme in the forms such as knowledge building-block, decision tree, artificial neural works, fuzzy system, graphs, etc., have also been proposed for efficient problem-solving. These meme-inspired algorithms, frameworks and paradigms have demonstrated with considerable success in various real-world applications.

Scope and Topics

The aim of this special session on memetic computing is to provide a forum for researchers in this field to exchange the latest advances in theories, technologies, and practice of memetic computing. The scope of this special session covers, but is not limited to:

  • Single/Multi-Objective memetic algorithms for continuous or combinatorial optimization.
  • Theoretical studies that enhance our understandings on the behaviors of memetic computing.
  • Adaptive systems and meme coordination.
  • Novel manifestations of memes for problem-solving.
  • Cognitive, Brain, individual learning, and social learning inspired memetic computation
  • Self-design algorithms in memetic computing.
  • Memetic frameworks using surrogate or approximation methods
  • Memetic automaton, cognitive and brain inspired agent based memetic computing
  • Data mining and knowledge learning in memetic computation paradigm
  • Memetic computing for expensive and complex real-world problems

CEC-24 Evolutionary Computation in Operations Research, Management Science and Decision Making

Organized by Wei-Chang Yeh and Yew-Soon Ong

Evolutionary Computation has been shown to attain high quality solutions to difficult optimization problems in fields for which exact and analytical methods do not perform well within tractable time, especially on big-scale problems, since the early 1990s. The essential idea of Evolutionary algorithms lies in the use of simple agents that work together in leading to emergent global behaviors that solve complex problems efficiently and effectively. In the recent years, there has been increasing interests to create new Evolutionary Computation methodologies by extending from existing Genetic algorithm (GA), Memetic Algorithm (MA), Ant Colony Optimization (ACO), Particle Swarm Optimization (PSO), Artificial Bee Colony (ABC) algorithms, Simplified Swarm Optimization (SSO), and others, that better emulates the power of nature in addressing big-scale real world problems in the field of Operations Research, Management Science and Decision Making. . The developed evolutionary algorithms are expected to be flexible to internal and external changes, robust even when some individuals fail, decentralized and self-organized.

In spite of the significant amount of research on Evolutionary Computation, there remain many open issues and intriguing challenges in addressing big-scale real world problems in the field of Operations Research, Management Science and Decision Making. The aims of this special session are to demonstrate the current state-of-the-art concepts of Evolutionary Computation in the field of Operations Research, Management Science and Decision Making, to reflect on the latest advances and showcase new directions in the area.

Scope and Topics

Authors are invited to submit their original and unpublished work in the areas including, but not limited to:

  • Evolutionary Computation
  • Advances in Evolutionary Computation
  • Evolutionary Computation applied to all fields of science and technology
  • Novel or Improved frameworks of Evolutionary Computation model
  • Knowledge incorporation in Evolutionary Computation,
  • Neural Networks
  • Fuzzy Systems
  • Multi-objective optimization
  • Robotics
  • Data Mining
  • Green logistic problems
  • Advanced transportation problems
  • Network design
  • Manufacturing cell design
  • Reliability design problems
  • Others

CEC-25 Bat Algorithm and Cuckoo Search

Organized by Xin-She Yang and Xingshi He

Nature-inspired algorithms, especially swarm intelligence based algorithms, have become very popular in optimization and computational intelligence. New algorithms such as the bat algorithm and cuckoo search have demonstrated some distinct advantages and have thus been applied in many areas such as engineering optimization and image segmentation. Though the literature about the practical applications has expanded significantly in the last few years, theoretical studies lack behind. [This special session will be the 2nd event of such topics, followed by the first successful event at CEC2015 in Japan.]

Scope and Topics

This special session intends to provide a timely platform to exchange ideas about new bio-inspired optimization algorithms, with emphasis on the following topics (but not limited to):

  • Theoretical analysis of new algorithms such as cuckoo search, firefly algorithm, and bat algorithm
  • Applications such as feature selection, classifications, design optimization and big data
  • Hybridization with other algorithms such as GA, PSO and others

CEC-26 Evolutionary Computation and Other Computational Intelligence Techniques for Cyber Security

Organized by Hongmei He

Internet of Things (IoT) delivers new value by connecting People, Process and Data. It brings great opportunities and is changing our life style. However, great opportunities also bring large risks that the IoT enabled systems could be threaten with various cyber-attacks, crimes and terrorism, as the IoT enabled systems produce a large cyber space. Hence, cyber security is particularly important in IoT enabled systems, and has raised much attention of researchers and industry recently. Evolutionary Computation and other Computational Intelligence techniques have been applied in various areas, such as computational biology, medical science, finance, engineering, etc. Cyber Security will be another area, where we can explore the power of Evolutional Computation and other Computational Intelligence techniques.

Scope and Topics

This special session will cover the following topics of Cyber Security enabled by Evolutionary Computation and other Computational Intelligence techniques, but not limited.

  • Bio-inspired cyber security architecture
  • Cloud security
  • Web spider defence
  • Biometrics-based authentication
  • Cyber-matrices based authentication
  • Prediction of attacks
  • Detection of spam emails
  • Detection and analysis of malware
  • Autonomous Security
  • Information security
  • Privacy and anonymity
  • Secure protocols

CEC-27 Evolutionary Computation in Dynamic and Uncertain Environments

Organized by Michalis Mavrovouniotis, Changhe Li, Shengxiang Yang and Yinan Guo

Many real-world optimization problems are subject to dynamism and uncertainties that are often impossible to avoid in practice. For instance, the fitness function is uncertain or noisy as a result of simulation/ measurement errors or approximation errors (in the case where surrogates are used in place of the computationally expensive high fidelity fitness function). In addition, the design variables or environmental conditions can be perturbed or they change over time.

The tools to solve these dynamic and uncertain optimization problems (DOP) should be flexible, able to tolerate uncertainties, fast to allow reaction to changes and adaptive. Moreover, the objective of such tools is no longer to simply locate the global optimum solution, but to continuously track the optimum in dynamic environments, or to find a robust solution that operates properly in the presence of uncertainties.

The last decade has witnessed increasing research efforts on handling dynamic and uncertain optimization problems using evolutionary algorithms and other metaheuristics, e.g., ant colony optimization, particle swarm optimization, artificial bee colony etc., and a variety of methods have been reported across a broad range of application backgrounds.

Scope and Topics

This special session aims at bringing together researchers from both academia and industry to review the latest advances and explore future directions in this field. Topics of interest include but are not limited to:

  • Benchmark problems and performance measures
  • Dynamic single - and multi-objective optimization
  • Adaptation, learning, and anticipation
  • Models of uncertainty and their management
  • Handling noisy fitness functions
  • Using fitness approximations
  • Searching for robust optimal solutions
  • Algorithm comparison and benchmarking
  • Hybrid approaches
  • Theoretical analysis
  • Real-world applications

CEC-28 Complex Networks and Evolutionary Computation

Organized by Jing Liu and Maoguo Gong

The application of complex networks to evolutionary computation (EC) has received considerable attention from the EC community in recent years. The most well‐known study should be the attempt of using complex networks, such as small‐world networks and scale‐free networks, as the potential population structures in evolutionary algorithms (EAs). Structured populations have been proposed to as a means for improving the search properties because several researchers have suggested that EAs populations might have structures endowed with spatial features, like many natural populations. Moreover, empirical results suggest that using structured populations is often beneficial owing to better diversity maintenance, formation of niches, and lower selection pressures in the population favouring the slow spreading of solutions and relieving premature convergence and stagnation. Moreover, the study of using complex networks to analyse fitness landscapes and designing predictive problem difficulty measures is also attracting increasing attentions. On the other hand, using EAs to solve problems related to complex networks, such as community detection, is also a popular topic.

Scope and Topics

This special session seeks to bring together the researchers from around the globe for a creative discussion on recent advances and challenges in combining complex networks and EAs. The special session will focus on, but not limited to, the following topics:

  • Complex networks and fitness landscape analysis
  • Complex networks and problem difficulty prediction
  • Evolutionary dynamics on complex networks
  • Evolutionary algorithms based on complex networks
  • Community detection using evolutionary algorithms
  • Community detection using multi‐objective evolutionary algorithms
  • Real world applications of evolutionary algorithms based on complex networks

CEC-29 Evolutionary Computation for Computational Biology

Organized by Anirban Mukhopadhyay, Ujjwal Maulik and Sanghamitra Bandyopadhyay

Computational biology is coming out as an emerging field for application of evolutionary computation tools and techniques such as genetic algorithms, genetic programming, differential evolution, particle swarm optimization, ant colony optimization and other related population-based metaheuristic techniques. Many of the computational biology problems, such as sequence alignment, gene mapping, fragment assembly, phylogenetic analysis, microarray analysis, biological network analysis and rational drug design can be posed as optimization problems. Therefore evolutionary computing techniques have been applied to these problems over the last few decades as optimization tools. However, growing size and complexity of biological data are creating new issues and challenges and it is becoming difficult to apply off-the-shelf techniques directly. These challenges include coping with large data size, handling many objective functions, dealing with large number of features, incorporating biological knowledge in the models etc. The main aim of this special session is to bring together the scientists and researchers of this field to exchange the latest advances in theories and experiments in this area.

Scope and Topics

Researchers are invited to submit original and unpublished works that deal with application of evolutionary computation techniques to the following and other related areas.

  • Sequence analysis including next-generation sequencing (sequence alignment, fragment assembly, gene mapping etc.).
  • Structure prediction (RNA and protein structure prediction, protein folding).
  • Microarray analysis (clustering, classification, feature selection etc.).
  • Genetic marker identification (cancer and other diseases).
  • Bio-molecule ordering and rank aggregation.
  • Protein-protein interaction prediction (intra-species and host-pathogen interactions).
  • Protein complex identification.
  • Protein sub-cellular location prediction.
  • Inferring gene regulatory and metabolic networks (including involvement of microRNAs).
  • Phyologenetic analysis and phylogenetic tree construction.
  • MicroRNA target prediction.
  • Drug target identification.
  • Differential network analysis and biological network alignment.
  • Biological motif finding (sequence and network motifs).
  • Rational drug design (molecular docking, ligand design etc.).
  • Multi-objective and many-objective optimization for computational biology problems.
  • Parallelization of evolutionary computing techniques for handling large biological data.

CEC-30 Optimization, Learning, and Decision-Making in Bioinformatics and Bioengineering (OLDBB)

Organized by Richard Allmendinger, Daniel Ashlock and Sanaz Mostaghim

Bioinformatics and Bioengineering (BB) are interdisciplinary scientific fields involving many branches of computer science, engineering, mathematics, and statistics. Bioinformatics is concerned with the development and application of computational methods for the modeling, retrieving and analysis of biological data, whilst Bioengineering is the application of engineering techniques to biology so as to create usable and economically viable products.

Bioinformatics and Bioengineering are relatively new fields in which many challenges and issues can be formulated as (single and multiobjective) optimization problems. These problems span from traditional problems, such as the optimization of biochemical processes, construction of gene regulatory networks, protein structure alignment and prediction, to more modern problems, such as directed evolution, drug design, experimental design, and optimization of manufacturing processes, material and equipment.

The main aim of this special session is to bring together both experts and new-comers working on Optimization and Decision-Making in Bioinformatics and Bioengineering (ODMBB) to discuss new and exciting issues in this area.

Scope and Topics

We encourage submission of papers describing new optimization strategies/challenges/applications/decision-making techniques in the area of BB. In addition, we are interested in application papers discussing the power and applicability of these novel methods to real-world problems in BB. You are invited to submit papers that are unpublished original work for this special session at IEEE CEC’16, which is part of the IEEE WCCI ‘16. The topics are, but not limited to, the following

  • (Single and multiobjective) optimization techniques for Bioinformatics and Bioengineering (BB) problems
    • Evolutionary algorithms
    • Swarm Intelligence
    • Metaheuristics
    • Fuzzy optimization
    • Hybrid optimization algorithms (combinations of heuristics and exact methods)
  • Decision-making techniques for BB problems
    • Preference elicitation and representation
    • Aggregation-based techniques
    • Fuzzy logic-based techniques
    • Bayesian-based techniques
  • Experimental optimization of BB problems
    • Experimental optimization platforms
    • Closed-loop optimization challenges and applications
    • Resourcing issues (interruptions, missing objective function values, changes of variables, etc.)
  • Learning in the optimization of BB problems
    • Link between Decision Maker’s learning and model’s learning
    • Capturing and learning from user preferences
    • Integrating optimization with machine learning
    • Interactive learning and optimization techniques
    • Techniques for learning user-driven parameter settings from examples
  • Tuning of optimization and decision-making techniques for BB problems
    • Performance measures
    • Test and benchmark problems
    • Visualization techniques
    • Optimization and visualization software
  • Emerging topics in BB
    • Novel applications (process design, manufacturing, etc)
    • Novel challenges (large-scale problems, dynamic problems, mixed integer problems, uncertainty, expensive and limited evaluations, etc)
    • Bilevel and multilevel optimization
    • Interactive visualization techniques
    • Multiobjective data mining
    • Predictive fitness landscape design
    • Many-objective optimization
    • Ecoinformatics
    • Side effect machines and other kernal representations for sequence analysis
    • Biomedical data modelling and mining

CEC-31 New Directions in Evolutionary Machine Learning

Organized by Will Browne, Keiki Takadama, Yusuke Nojima, Masaya Nakata and Tim Kovacs

Evolutionary Machine Learning (EML) explores technologies that integrate machine learning with evolutionary computation for tasks including optimization, classification, regression, and clustering. Since machine learning contributes to a local search while evolutionary computation contributes to a global search, one of the fundamental interests in EML is a management of interactions between learning and evolution to produce a system performance that cannot be achieved by either of these approaches alone. Historically, this research area was called GBML (genetics-based machine learning) and it was concerned with learning classifier systems (LCS) with its numerous implementations such as fuzzy learning classifier systems (Fuzzy LCS). More recently, EML has emerged as a more general field than GBML; EML covers a wider range of machine learning adapted methods such as genetic programming for ML, evolving ensembles, evolving neural networks, and genetic fuzzy systems; in short, any combination of evolution and machine learning. EML is consequently a broader, more flexible and more capable paradigm than GBML. From this viewpoint, the aim of this special session is to explore potential EML technologies and clarify new directions for EML to show its prospects.

Scope and Topics

This special session follows the first successful special session (the largest session among the special sessions) held in CEC 2015. The continuous exploration of this field by organizing the special session in CEC is indispensable to establish the discipline of EML. For this purpose, this special session focuses on, but is not limited to, the following areas in EML:

  • Evolutionary learning systems (e.g., learning classifier systems)
  • Evolutionary fuzzy systems
  • Evolutionary reinforcement learning
  • Evolutionary neural networks
  • Evolutionary adaptive systems
  • Artificial immune systems
  • Genetic programming applied to machine learning
  • Transfer learning; learning blocks of knowledge (memes, code, etc.) and evolving the sharing to related problem domains
  • Accuracy-Interpretability tradeoff in EML
  • Applications and theory of EML

CEC-32 Differential Evolution: Past, Present and Future

Organized by Kai Qin, Kenneth V. Price, Swagatam Das and Jouni Lampinen

Differential evolution (DE) emerged as a simple and powerful stochastic real-parameter optimizer more than a decade ago and has now developed into one of the most promising research areas in the field of evolutionary computation. The success of DE has been ubiquitously evidenced in various problem domains, e.g., continuous, combinatorial, mixed continuous-discrete, single-objective, multi-objective, constrained, large-scale, multimodal, dynamic and uncertain optimization problems. Furthermore, the remarkable efficacy of DE in real-world applications significantly boosts its popularity.

Over the past decades, numerous studies on DE have been carried out to improve the performance of DE, to give a theoretical explanation of the behavior of DE, to apply DE and its derivatives to solve various scientific and engineering problems, as demonstrated by a huge number of research publications on DE in the forms of monographs, edited volumes and archival articles. Consequently, DE related algorithms have frequently demonstrated superior performance in challenging tasks. It is worth noting that DE has always been one of the top performers in previous competitions held at the IEEE Congress on Evolutionary Computation. Nonetheless, the lack of systematic benchmarking of the DE related algorithms in different problem domains, the existence of many open problems in DE, and the emergence of new application areas call for an in-depth investigation of DE.

Scope and Topics

This special session aims at bringing together researchers and practitioners to review and re-analyze past achievements, to report and discuss latest advances, and to explore and propose future directions in this rapidly emerging research area. Authors are invited to submit their original and unpublished work in the areas including, but not limited to:

  • DE for continuous, discrete, mixed, single-objective, multi-objective, constrained, large-scale,
  • multiple optima seeking, dynamic and uncertain optimization
  • Review, comparison and analysis of DE in different problem domains
  • Experimental design and empirical analysis of DE
  • Study on initialization, reproduction and selection strategies in DE
  • Study on control parameters (e.g. scale factor, crossover rate, population size) in DE
  • Self-adaptive and tuning-free DE
  • Parallel and distributed DE
  • Theory of DE
  • Synergy between DE and machine learning techniques
  • Hybridization of DE with other optimization techniques
  • Application of DE to real-world problems

CEC-33 Evolutionary Algorithms for Mixed-Integer Optimization Problems

Organized by Martin Schlueter, Hernan Aguirre and Akira Oyama

Many real-world applications are based on both: continuous and discrete parameters. Optimization models that consider these two kind of parameters simultaneously are referred to as mixed-integer problems and are exceptionally difficult to solve. While there exists comprehensive analysis of deterministic algorithms for mixed-integer problems, evolutionary algorithms for this kind of problem is still a young and emerging field. Considering the robustness of evolutionary algorithms and their often existing capability for parallelization and multi/many-objective optimization, evolutionary algorithms can offer a significant new potential for the class of mixed-integer problems. In the context of evolutionary computing, this will be the first session especially dedicated to mixed-integer problems.

Scope and Topics

This special session intends to bring together researchers who apply evolutionary algorithms and other search heuristics on mixed integer optimization problems. Aim of this session is to provide a forum, where experience and new techniques in solving mixed-integer problems via such methods is shared and discussed. The main topics include but are not limited to:

  • Evolutionary algorithms and other search heuristics
  • Hybrid approaches combining various methods
  • Computational comparisons
  • Theoretical analysis
  • Single and multi/many-objective problems
  • Constraint and unconstrained optimization
  • Real-world applications
  • Academic benchmark sets
  • Parallelization techniques for mixed-integer algorithms
  • Large-scale mixed integer problems
  • Mixed-integer problems with stochastic noise

CEC-34 Evolutionary Multi-objective Optimization

Organized by Sanaz Mostaghim and Kalyanmoy Deb

This special session invites papers discussing recent advances in the development and application of biologically-inspired multi-objective optimization algorithms.

Many problems from science and industry have several (and normally conflicting) objectives that have to be optimized at the same time. Such problems are called multi-objective optimization problems and have been subject of research in the past two decades. One of the reasons why evolutionary algorithms are so suitable for multi-objective optimization is because they can generate a whole set of solutions (the Pareto-optimal solutions) in a single run rather than requiring an iterative one-solution-at-a-time process as followed in traditional mathematical programming techniques.

The main aim of this special session organized within the 2016 IEEE Congress on Evolutionary Computation (CEC'2016) is to bring together both experts and new-comers working on Evolutionary Multi-objective Optimization (EMO) to discuss new and exciting issues in this area.

Scope and Topics

We encourage submission of papers describing new concepts and strategies, and systems and tools providing practical implementations, including hardware and software aspects. In addition, we are interested in application papers discussing the power and applicability of these novel methods to real-world problems in different areas in science and industry. You are invited to submit papers that are unpublished original work for this special session at CEC 2015. The topics are, but not limited to, the following

  • Theoretical aspects of EMO algorithms
  • Real-world applications of EMO algorithms
  • Test and benchmark problems for EMO algorithms
  • Multi-objectivization and visualization techniques
  • Innovization
  • New EMO techniques including those using meta-heuristics such as artificial immune systems, particle swarm optimization, differential evolution, cultural algorithms, etc.
  • Handling practicalities, such as uncertainty, noise, constraints, dynamically changing problems, bi-level problems, mixed-integer problems, computationally expensive problems, fixed budget of evaluations, etc.
  • Performance measures for EMO algorithms
  • Techniques to keep diversity in the population
  • Comparative studies of EMO algorithms
  • Evolutionary multi-objective combinatorial optimization, EMO control problems, EMO inverse problems, EMO data mining, EMO machine learning
  • Memetic and Metaheuristics based EMO algorithms
  • Hybrid approaches combining, for example, EMO algorithms with mathematical programming techniques and exact methods
  • Parallel EMO approaches
  • Many-objective optimization
  • Adaptation, learning, and anticipation

CEC-35 Parallel and Distributed Evolutionary Computation in the Inter-Cloud Era

Organized by Masaharu Munetomo, Juan Julián Merelo Guervós and Yuji Sato

Recent advances in cloud computing lead to a global infrastructure of “the Inter-cloud” (clouds of cloud systems) that can be utilized through the Internet to provide with virtually infinite IT resources such as virtual machines and storage units just by calling web-service APIs through the Internet.

It is necessary to have enough resources and complexities in the environment for the individuals to “evolve”. Cloud systems may even offer tens of thousands of virtual machines, terabytes of memories and exabytes of storage capacity. Current trend toward many-core architecture increases the number of cores even more dramatically: we may have more than a million of cores to offer extremely massive parallelization.

In this special session, we will discuss parallel and distributed evolutionary computation in the cloud era such as implementation of massively parallel evolutionary algorithms employing cloud computing systems and services, parallel implementation of evolutionary algorithms on many-core architectures including GPUs, and we also welcome any types of parallel and distributed evolutionary computation on any “informal” types of computing environment in this special session including the following themes.

Scope and Topics

The topics are, but not limited to, the following

  • Implementation of parallel and distributed evolutionary computation in cloud computing systems and/or services
  • Implementation of massively parallel evolutionary computation on many-core architecture such as GPUs
  • Parallel and distributed evolutionary machine learning techniques
  • Design and theory of scalable evolutionary algorithms
  • Development of parallel and distributed evolutionary computation framework in cloud computing systems
  • Applications of parallel and evolutionary computation techniques in cloud or other modern computing environment
  • Applications of EC and other bio-inspired paradigms to peer to peer systems, and distributed EC algorithms that use them.
  • Peer-to-peer computing, volunteer computing and zero-cost distributed computing. Large scale autonomous systems, sneaky or parasite computing using the browser or other widely available infrastructure. Internet of Things or Everything (IoT or IoE).

CEC-36 Evolutionary Computation in Architectural Design

Organized by Fatih Tasgetiren, Sevil Sariyildiz, Ozer Ciftcioglu and Suganthan

Architecture has a profound impact on the environment that we spend much of our daily lives in. As such, it is important that the outcomes of architectural design are well performing and suitable for their intended purpose. Architectural design is a process of high complexity, which aims to satisfy design goals comprising hard, engineering aspects, together with soft, perceptual and cognitive ones. Due to the excessive complexity of the design task, human cognition alone is often insufficient to ensure suitable outcomes. Computational Intelligence and Soft Computing methods can aid in confidently arriving at high-performing architectural design solutions, as well as provide valuable inspiration during the design process. As such, these methods are high on the contemporary scientific agenda.

The majority of architectural design problems involve real-valued decision variables and multiple objectives. For these reasons, they prove to be an ideal field for Multi-Objective Real-Parameter Constrained Optimization applications of soft computing and computational intelligence methods such as Evolutionary Algorithms, machine learning, fuzzy logic, simulation etc. These methods are capable of navigating complex design spaces, such as those encountered in architectural design problems, and identifying best tradeoff solutions.

Scope and Topics

This session aims to put forward original contributions, latest research and development, and contemporary issues in the field of soft computing and computational intelligence for architectural and building design. It intends to collect a series of innovative, high quality papers on ideas, concepts, and technologies that make use of evolutionary algorithms in these research areas. Proposed submissions should be original, unpublished, and present novel fundamental research contributions from a theoretical or an application point of view. Session topics include (but are not limited to) the following:

  • Evolutionary Computation and Multi-Objective Optimization in Architectural and Urban Design.
  • Surrogate Modelling and Meta-models in Sustainable Architectural Design
  • Modeling of Fuzzy and Imprecise Design Aspects
  • Soft Computing in Spatial, Formal and Color Perception
  • Cognitive Models for Supporting Design Decisions
  • Applications in Practice: Urban Design
  • Applications in Practice: Large Scale Structures and Complexes
  • Soft Computing and Architectural education

CEC-37 Theoretical Foundations of Bio-inspired Computation

Organized by Pietro S. Oliveto and Andrew M. Sutton

Bio-inspired search heuristics often turn out to be highly successful for optimization in practice. The theory of these randomized search heuristics explains the success or the failure of these methods in practical applications. Theoretical analyses lead to the understanding of which problems are optimized (or approximated) efficiently by a given algorithm and which are not. The benefits of theoretical understanding for practitioners are threefold.

  1. Aiding the algorithm design,
  2. guiding the choice of the best algorithm for the problem at hand,
  3. determining the optimal parameter settings.
The theory of evolutionary computation has grown rapidly in recent years. The primary aim of this special session is to bring together people working on theoretical aspects of bio- inspired computation. The latest breakthroughs in the theory of bio-inspired computation will be reported and new directions will be set.

Scope and Topics

Potential authors are invited to submit papers describing original contributions to foundations of evolutionary computation. Although we are most interested in theoretical foundations, computational studies of a foundational nature are also welcome. The scope of this special session includes (but is not limited to) the following topics:

  • Theoretical foundations of bio-inspired heuristics
  • Exact and approximation runtime analysis
  • Black box complexity
  • Self-adaptation
  • Population dynamics
  • Fitness landscape and problem difficulty analysis
  • No free lunch theorems
  • Statistical approaches for understanding the behavior of bio-inspired heuristics
  • Computational studies of a foundational nature
All problem domains will be considered including:
  • combinatorial and continuous optimization
  • single-objective and multi-objective optimization
  • constraint handling
  • dynamic and stochastic optimization
  • co-evolution and evolutionary learning

CEC-38 Evolutionary computation with human factors

Organized by Yan Pei and Hideyuki Takagi

Evolutionary computation (EC) with human factors, such as Interactive EC (IEC), EC for human-related applications, EC-based visual/auditory/haptic design, EC for analyzing human characteristics and others, is an approach whereby such properties as human knowledge, experience, and preference are embedded into an optimization/design process and application tasks. By embedding the human being itself into an optimization system or using EC for humans, EC techniques become applicable to tasks for which it is difficult to construct an evaluation system or measure the evaluations. EC with human factors has also been applied to artistic areas such as creating music or graphics, engineering areas such as sound and image processing, control and robotics, virtual reality, data mining, media database retrieval, and others, including geoscience, education, games, and many other tasks in various areas. From a framework point of view, EC with human factors can be realized with any EC algorithm by replacing the fitness function with a human user. Several EC techniques are used in EC with human factors, such as interactive genetic algorithms, interactive genetic programming, interactive evolution strategy, human based genetic algorithm, interactive particle swarm optimization, interactive differential evolution, and others. There are many directions in EC with human factors research, such as expanding the applications of EC with human factors; expanding EC with human factors frameworks; applying EC with human factors in a reverse engineering approach to analyze humans and thus advance; accelerating EC with human factors searches and improving IEC interfaces. The proposed session provides researchers and educators a platform to report and discuss state of the art research and study in subjects of evolutionary computation with human factors.

Scope and Topics

The topics within (but not limited in) these scopes are especially welcome to submit to this special session:

  • IEC algorithm development and application, such as interactive genetic algorithm,interactive genetic programming, interactive differential evolution, Human based-genetic algorithm, and interactive particle swarm optimization
  • New IEC algorithms and EC with human factors frameworks
  • EC for human science and Kansei engineering
  • EC for analyzing human characteristics
  • EC-based visual/auditory/haptic design
  • Human centering computing with EC
  • Human model study and development
  • Fitness landscape study in EC with human factors
  • Human computer cooperation with evolutionary computation
  • Applications of EC with human factor techniques
  • Theoretical fundamental of EC with human factors

CEC-39 Transfer Learning in Evolutionary Computation

Organized by Mengjie Zhang, Muhammad Iqbal, Yi Mei and Bing Xue

Data mining, machine learning, and optimisation algorithms have achieved promises in many real-world tasks, such as classification, clustering and regression. These algorithms can often generalise well on data in the same domain, i.e. drawn from the same feature space and with the same distribution. However, in many real-world applications, the available data are often from different domains. For example, we may need to perform classification in one target domain, but only have sufficient training data in another (source) domain, which may be in a different feature space or follow a different data distribution. Transfer Learning aims to transfer knowledge acquired in one problem domain, i.e. the source domain, onto another domain, i.e. the target domain. Transfer learning has recently emerged as a new learning framework and hot topic in data mining and machine learning.

Scope and Topics

Evolutionary computation techniques have been successfully applied to many real-world problems, and started to be used to solve transfer learning tasks. Meanwhile, transfer learning has attracted increasing attention from many disciplines, and has been used in evolutionary computation to address complex and challenging issues. The theme of this special session is transfer learning in evolutionary computation, covering ALL different evolutionary computation paradigms. The aim is to investigate in both the new theories and methods on how transfer learning can be achieved with different evolutionary computation paradigms, and how transfer learning can be adopted in evolutionary computation, and the applications of evolutionary computation and transfer learning in real-world problems. Authors are invited to submit their original and unpublished work to this special session. Topics of interest include but are not limited to: 

  • Evolutionary supervised transfer learning
  • Evolutionary unsupervised transfer learning
  • Evolutionary semi-supervised transfer learning
  • Domain adaptation and domain generalization in evolutionary computation
  • Instance based transfer approaches in evolutionary computation
  • Feature based transfer learning in evolutionary computation
  • Parameter/model based transfer learning in evolutionary computation
  • Relational based transfer learning in evolutionary computation
  • Transfer learning in in evolutionary computation for classification
  • Transfer learning in in evolutionary computation for regression
  • Transfer learning in in evolutionary computation for clustering
  • Transfer learning in in evolutionary computation for other data mining tasks, such as association rules and link analysis
  • Transfer learning in in evolutionary computation for scheduling and combinatorial optimisation tasks
  • Hybridisation of evolutionary computation and neural networks, and fuzzy systems for transfer learning
  • Hybridisation of evolutionary computation and machine learning, information theory, statistics, etc., for transfer learning
  • Transfer learning in in evolutionary computation for real-world applications, e.g. text mining, image analysis, face recognition, WiFi localisation, et al.

CEC-40 Advances in Decomposition­based Evolutionary Multi­objective Optimization (ADEMO)

Organized by Saúl Zapotecas­-Martínez, Bilel Derbel, Qingfu Zhang and Carlos Artemio Coello Coello

The purpose of this special session is to promote the design, study, and validation of generic approaches for solving multi­objective optimization problems based on the concept of decomposition. Decomposition­based Evolutionary Multi­objective Optimization (DEMO) encompasses any technique, concept or framework that takes inspiration from the “divide and conquer” paradigm, by essentially breaking a multi­objective optimization problem into several sub­problems for which solutions for the original global problem are computed and aggregated in a cooperative manner. This simple idea, which is rather standard in computer science and information systems, allows to open up new exciting research perspectives and challenges both at the fundamental level of our understanding of multi­objective problems, and in terms of designing and implementing new efficient algorithms for solving them. Generally speaking, the special session will focus on stochastic evolutionary approaches for which decomposition is performed with respect to the objective space, typically by means of scalarizing functions like in the MOEA/D framework. We, however, encourage contributions reporting advances with respect to other decomposition techniques operating in the decision space as done in the co­called cone­speration methods; or other hybrid approaches taking inspiration from operations research and mathematical programming. In fact, many different DMOEAs variants have been proposed, studied and applied to various application domains in recent years. However, DMOEAs are still in their very early infancy, since only few basic design principles have been established compared to the huge body of literature dedicated to other well­established approaches (e.g. Pareto ranking, indicator­based techniques, etc), and relatively few research forums have been dedicated to the study of DEMO approaches and their unification. The main goal of the proposed session is to encourage research studies that systematically investigate the critical issues in DMOEAs at the aim of understanding their key ingredients and their main dynamics, as well a to develop solid and generic principles for designing them. The long term goal is to contribute to the emergence of a general and unified methodology for the design, the tuning and the performance assessment of DMOEAs.

Scope and Topics

The special session will be a nice opportunity for researchers in the evolutionary and multiobjective optimization filed to exchange their recent ideas and advances on the design and analysis of DEMO approaches. In this respect, we are welcoming high quality papers in theoretical, developmental, implementational, and applied aspects of DEMO approaches. More particularly, the special session will encourage original research contributions that address new and existing DMOEAs, their contributions and relationships to other methodologies dedicated to multi­objective optimization in terms of: algorithmic components, decomposition strategies, collaboration among different search procedures, design of new specialized search procedures, parallel and distributed implementations, incorporation of user interaction, combination and hybridization with other traditional (heuristic or exact) techniques, strategies for dealing with many objectives, noisy problems, and expensive problems, problem solving and applications, etc. The main focus will be on eliciting the main design principles that lead to effective and efficient cooperative search procedures among the so­ defined single or multiple objective subproblems.

The topics of interests include (but are not limited to) the following issues :

  • Analysis of algorithmic components and performance assessment of DEMO approaches
  • Experimental and theoretical investigations on the accuracy of the underlying decomposition strategies, e.g. scalarizing functions techniques, multiple reference points, variable grouping, etc.
  • Adaptive, self­adaptive, and tuning aspects for the parameter setting and configuration of DEMO approaches
  • Design and analysis of new DEMO approaches dedicated to specific combinatorial, constrained and/or continuous domains
  • Effective hybridization of single­objective solvers with DEMO approaches, i.e., plug­and­ play algorithms based on traditional single objective evolutionary algorithms and meta­ heuristics, such as: Genetic Algorithms (GAs), Particle Swarm Optimization (PSO), Differential Evolution (DE), Ant Colony Optimization (ACO), Covariance Matrix Evolution Strategy (CMA­ES), Scatter Search (SS), etc.
  • Adaptation and analysis of DEMO approaches in the context of large scale and many objective problem solving
  • Application of DEMO for real­world problem solving
  • Design and implementation of DEMO approaches in massively parallel and large scale distributed environment (e.g., GPUs, Clusters, Grids, etc)
  • Software tools for the design implementation and performance assessment of DEMO approaches

CEC-41 Niching Methods for Multimodal Optimization

Organized by Michael G. Epitropakis, Xiaodong Li and Andries Engelbrecht

Population or single solution search­based optimization algorithms (i.e. {meta,hyper}­heuristics) in their original forms are usually designed for locating a single global solution. Representative examples include among others evolutionary and swarm intelligence algorithms. These search algorithms typically converge to a single solution because of the global selection scheme used. Nevertheless, many real­world problems are “multimodal” by nature, i.e., multiple satisfactory solutions exist. It may be desirable to locate many such satisfactory solutions, or even all of them, so that a decision maker can choose one that is most proper in his/her problem domain. Numerous techniques have been developed in the past for locating multiple optima (global and/or local). These techniques are commonly referred to as “niching” methods. A niching method can be incorporated into a standard search­based optimization algorithm, in a sequential or parallel way, with an aim to locate multiple globally optimal or suboptimal solutions,. Sequential approaches locate optimal solutions progressively over time, while parallel approaches promote and maintain formation of multiple stable subpopulations within a single population. Many niching methods have been developed in the past, including crowding, fitness sharing, derating, restricted tournament selection, clearing, speciation, etc. In more recent times, niching methods have also been developed for meta­heuristic algorithms such as Particle Swarm Optimization, Differential Evolution and Evolution Strategies.

Scope and Topics

Most of existing niching methods, however, have difficulties that need to be overcome before they can be applied successfully to real­world multimodal problems. Some identified issues include: difficulties to pre­specify some niching parameters; difficulties in maintaining found solutions in a run; extra computational overhead; poor scalability when dimensionality and modality are high. This special session aims to highlight the latest developments in niching methods, bringing together researchers from academia and industries, and exploring future research directions on this topic. We invite authors to submit original and unpublished work on niching methods. Topics of interest include but are not limited to:

  • Theoretical analysis of niching methods
  • Handling the issue of niching parameters in niching methods
  • Niching methods that incurs lower computational costs
  • Handling the scalability issue (both dimensionality and modality) in niching methods
  • Adaptive or parameter­less niching methods
  • Multiobjective approaches to niching
  • Multimodal optimization in dynamic environments
  • Niching methods applied to discrete optimization problems
  • Niching methods applied to constrained optimization problems
  • Benchmarking niching methods, including test functions and performance metrics
  • Comparative studies of various niching methods
  • Niching methods applied to engineering and other real­world optimization problems
Please note that we are NOT interested if the adopted task is to find a single solution of a multimodal problem.

Furthermore, a companion Competition on Niching Methods for Multimodal Optimization (subject acceptance​ be organized in conjunction with this special session. The aim of the) will competition is to provide a common platform that encourages fair and easy comparisons across different niching algorithms. The competition allows participants to run their own niching algorithms on 20 benchmark multimodal functions with different characteristics and levels of difficulty. Detailed information about these benchmark functions is provided in the following technical report:
  • X. Li, A. Engelbrecht, and M.G. Epitropakis, ``Benchmark Functions for CEC'2013 Special Session and Competition on Niching Methods for Multimodal Function Optimization'', Technical Report, Evolutionary Computation and Machine Learning Group, RMIT University, Australia, 2013.
Researchers are welcome to evaluate their niching algorithms using this benchmark suite, and report the results by submitting a paper to the associated niching special session (i.e., submitting via the online submission system of CEC’2016). In case it is too late to submit the paper (i.e., passing the CEC'2016 submission deadline), author may submit their results in a report directly to the special session organizers, in order to be counted in the competition.

CEC-42 Special Session Associated with Competition on Bound Constrained Single Objective Numerical Optimization

Organized by Suganthan, Mostafa Z. Ali, Qin Chen, Liang and B. Y. Qu

This special session will be used to receive conference papers submitted to the competition on the above competition. The details of the competition are presented below. These details are equally applicable to the SS also.

Competition Goals

The goals are to evaluate the current state of the art in single objective optimization with bound constraints and to propose novel benchmark problems with diverse characteristics. The algorithms will be evaluated with very small number of function evaluations to large number of function evaluations as well as single solution to multiple solutions. Under the above scenarios, novel problems will be designed for the first time to emulate real-world problem solving. In particular the following cases would also be considered:

  1. An industry is interested in solving different instances of the same class of problem. How can we learn from past instances to solve the future instances more effectively?
  2. For a given single objective problem, locate N top solutions separated from each other by a specified Euclidean distance. These two cases have not been considered much in the evolutionary computation community, although these scenarios are commonly encountered in real-world problem solving settings.

Contributions to the Evolutionary Computation Community

Single objective numerical optimization is the most important class of problems. All new evolutionary and swarm algorithms are tested on single objective benchmark problems. In addition, these single objective benchmark problems can be transformed into dynamic, niching composition, computationally expensive and many other classes of problems.

How to submit an entry and how to evaluate them

Potential authors are asked to make use of the codes of benchmark problems to be distributed from the competitions web pages to test their algorithms either with or without surrogate methods. The authors have to execute their novel or existing algorithms on the given benchmark problems and present the results in various formats as outlined in the technical report. The evaluation criteria will also be specified in the technical report. The authors are asked to prepare a conference paper detailing the algorithms used and the results obtained on the given benchmark problems and submit their papers to the associated special session within CEC 2016. The authors presenting the best results should also be willing to release their codes for verification before declaring the eventual winners of the competition.

Special Session Associated With This Competition

This competition requires all entries to have an associated conference paper submitted. We also expect at least one author of each entry to register, attend the conference and present their papers.

CEC-43 ​Evolutionary Physical Systems and Matter

Organized by Stefano Nichele and Gunnar Tufte

The special session on “evolution of physical systems and matter” encompasses understanding, modeling and applying biologically inspired mechanisms to physical systems, where evolution occurs entirely in the real­world physical substrates rather than in simulation. Using real physical systems or materials for computation may allow evolution to exploit underlying physical properties that may not be available in simulation (e.g. due to «reality gap»), thus allowing the discovery of novel evolutionary solutions.
The aim of this special session is to bring together researchers in order to share ideas and innovations on biologically inspired mechanisms applied to physical systems. Application areas include bio­inspired algorithms applied to physical systems, the creation of novel physical devices, novel or optimized designs for physical systems, adaptive physical systems, novel evolutionary techniques for embedded evolution and embedded computation, and novel material substrates that may support computation.
This special session is inspired by overlapping principles that emerged in several domains, such as Evolution­in­Materio (Pask, 1959; Miller and Downing, 2002), where underlying physics of materials is used as computation substrate.
Notable examples of evolution of physical systems and matter range from novel nanoscale materials for computation (Broersma et al., 2012), to FPGAs (Thompson, 1996), embodied evolution (Watson et al., 2002), 3D printers (Rieffel and Sayles, 2010), EHW (Yao and Higuchi, 1996; Greenwood and Tyrrell, 2006), electronic devices (Hornby at al., 2006) and robotics (Zykov et al., 2004).

Scope and Topics

The special session on Evolutionary Physical Systems and Matter intends to collect both theoretical machines and contributions/principles, and practical applications. Real application scenarios, from nanotechnology to buildings, are welcome. The topics of this special issue include (but are not limited to):

  • Evolution of Physical Systems
  • Evolution­in­Materio
  • Computational Matter
  • Evolution of Micro/Nano­Devices
  • Embodied Evolution
  • Morphological Computation (e.g. Tensegrity, 3D printers) and Embodied Computation
  • Slime Mould Computing
  • Physical Implementations of Reservoir Computing
  • Exploitation of Evolutionary Techniques for Creating Physical Circuits (Design and Fabrication)
  • Micro and Nano­scale Electronic Chemistry
  • Evolvable Systems Techniques
  • Self­reconfiguration, fault tolerance, self repair, adaptation
  • Artificial Generative and Developmental Systems
  • Evolution / Co­evolution of Controllers and Morphologies
  • Bio­Inspired Computation for new Materials Engineering
  • Bio­Inspired Computation for Fabrication of Novel Physical Devices

CEC-44 Large Scale Global Optimization

Organized by Daniel Molina, Swagatam Das and Antonio La Torre

In the past two decades, many nature-inspired optimization algorithms have been developed and applied successfully for solving a wide range of optimization problems, including Simulated Annealing (SA), Evolutionary Algorithms (EAs), Differential Evolution (DE), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Artificial Bee Colony (ABC), Estimation of Distribution Algorithms (EDA), etc. Although these techniques have shown excellent search capabilities when applied to small or medium sized problems, they still encounter serious challenges when applied to large scale problems, i.e., problems with several hundreds to thousands of variables. The reasons appear to be two-fold. Firstly, the complexity of a problem usually increases with the increasing number of decision variables, constraints, or objectives (for multi-objective optimization problems), which may prevent a previously successful search strategy from locating the optimal solutions. Secondly, as the size of the solution space of the problem grows exponentially with the increasing number of decision variables, there is an urgent need to develop more effective and efficient search strategies to better explore this vast solution space with a limited computational budget.

In recent years, researches on scaling up EAs to large scale problems have attracted much attention, including both theoretical and practical studies. Existing work on this topic are still however rather limited, given the significance of the scalability issue.

Scope and Topics

This special session is devoted to highlight the recent advances in EAs for handling large scale global optimization (LSGO) problems, involving single or multiple objectives problems, unconstrained or constrained search spaces and binary/discrete, real, or mixed decision variables. More specifically, we encourage interested researchers to submit their original and unpublished work on:

  • Theoretical and experimental analysis of the scalability of EAs.
  • Novel approaches and algorithms for scaling up EAs to large scale optimization problems.
  • Applications of EAs to real-world large scale optimization problems.
  • Specialized algorithms for large scale optimization.
  • Novel test suites that help us to understand large scale optimization problem characteristics.
  • New grouping approaches for cooperative coevolution.


  • There is not proposed a particular competition, however, for people interested in LSGO for real-coding, we recommend the following benchmark using a dimensionality of 1000 variables:

    X. Li, K. Tang, M. Omidvar, Z. Yang and K. Qin, "Benchmark Functions for the CEC'2013 Special Session and Competition on Large Scale Global Optimization," Technical Report, Evolutionary Computation and Machine Learning Group, RMIT University, Australia, 2013

CEC-45 Evolutionary Computing Application in Hardware

Organized by Andy M Tyrrell and Martin A Trefzer

Evolvable systems encompass understanding, modelling and applying biologically inspired mechanisms to physical systems. Application areas for bio-inspired algorithms include the creation of novel physical devices/systems, novel or optimised designs for physical systems and for the achievement of adaptive physical systems. Having showcased examples from analogue and digital electronics, antennas, MEMS chips, optical systems as well as quantum circuits in the past, we are looking for papers that apply techniques and applications of evolvable systems to these hardware systems, and in particular this year looking for papers in the areas of evolutionary robotic and evolutionary many-core system.

Scope and Topics

Topics include but are not limited to:

  • Evolvable Systems Techniques:
    • Self-reconfigurable systems
    • Adaptive systems
    • Self-repeating systems
    • Fault tolerant systems
    • Autonomous systems
    • Specialised hardware
  • Evolutionary Robotics:
    • Learning and adaptation
    • Cooperation and competition
    • Co-evolution of robot morphologies
    • Real-world applications
  • Evolutionary Many-core Systems:
    • Self-adaptation
    • Self-monitoring
    • Self-testing
    • Hardware system optimisation

CEC-46 Bio-inspired and Evolutionary Computation methods for unsupervised learning

Organized by David Camacho

This special session will be devoted to the study of new Bio-inspired and Evolutionary-based algorithms which have been applied into open, dynamic and complex problems where the application of Unsupervised Learning algorithms can provide new insights in these kind of domains. The special session will be particularly focused on the practical application of bio-inspired methods (evolutionary strategies, swarm intelligence methods, or nature-based approaches amongst others) and their application to unsupervised algorithms as Clustering (K-means, Hierarchical clustering, etc.), Grouping, Hidden Markov Models, or Latent variable models (EM algorithm, method of moments, …) approaches to mention only some few. The application to real and complex scenarios, as community finding in very large social networks, data analysis in wireless sensor networks, real applications in industry and engineering, automatic and semi-automatic malware detection, energy, and so many others, will be welcomed. Therefore, the main goals of this special session will be two-fold: On the one hand, to look for new algorithms and techniques proposals based on Bio-inspired and EC, which have been successfully combined with other unsupervised approaches. On the other hand, to look for new application domains, and real problems, where the application of Bio-inspired and EC combined with unsupervised methods have demonstrated an outstanding performance against other traditional approaches.

Paper acceptance and publication will be judged on the basis of their quality and relevance to the symposium themes, clarity of presentation, originality and accuracy of results and proposed solutions.

Scope and Topics

Topics include, but are not limited to:

  • Theoretical models for Bio-inspired (evolutionary strategies, swarm intelligence algorithms, nature-based methods) approaches to unsupervised algorithms (grouping, clustering, HMM, Latent-based, etc.).
  • Application of Evolutionary and Bio-inspired methods in Grouping and Clustering.
  • Application of Evolutionary and Bio-inspired methods in HMM-based approaches.
  • Application of Evolutionary and Bio-inspired methods for Latent-based models (factor analysis, PCA, ICA, mixture of gaussians,…)
  • Hybridization and Application of Swarm Intelligence (ACO, PSO,…) methods in grouping, graph-based, clustering, HMM, statistical and other unsupervised learning algorithms.
  • Bio-inspired and Evolutionary Computation strategies for Unsupervised Learning methods focused on semantic applications such as automatic summarization, recommender systems, topic identification, etc.
  • Bio-inspired and EC Computation methods for Unsupervised Learning in real problems: social-based mining, counter-terrorism, energy, wireless sensor networks, malware detection and classification, videogames, unmanned systems, e-business/e-commerce, e-learning, e-health, e-science, e-government, crisis management, etc.

CEC-47 Quantum Computing and Evolutionary Computation

Organized by Martin Lukac, William N. N. Hung and Claudio Moraga

As quantum information and computation research continues to develop, we will see increasing interest in adapting the philosophy of quantum computing, information theory and ideology into other, more traditional aspects of computational research. Although the hardware technology to realize quantum computing is still yet to be materialized, research about the theoretical aspects of quantum computing and its ideology has enjoyed some success with artificial and computational intelligence. The main aim of this special session is to bring together experts in algorithms, quantum information, quantum algorithms, physicists and hardware designers to discover new applications and features resulting from not only a cross-disciplinary information exchange but also from discussion between engineers and scientists working in the larger area of quantum information and computation.

Scope and Topics

This special session focus on combining various aspects of quantum computing, information theory, and other aspects with existing fields in computational intelligence. In particular the usage of classical algorithms to solve quantum problems, usage of quantum algorithms to solve classical problems and the usage of quantum algorithms for solving quantum problems are three main sought general ideas to be the center focus of this special session. Some typical research areas that will be discussed in this special session include (but are not limited to) the following:

  • Quantum inspired evolutionary computation, quantum inspired genetic algorithms.
  • Quantum evolutionary computing related areas such as quantum neural networks and quantum fuzzy computing systems.
  • Evolutionary Techniques and Quantum Computing. Including: (a) use of evolutionary paradigms to create quantum circuits, quantum algorithms, quantum architectures and quantum games, (b) creation of new quantum algorithms and architectures inspired by the concepts of evolution and other biological ideas,(c) use of evolutionary algorithms to solve any practical problems in designing quantum devices.
  • Quantum implementation of Computational Intelligence: many machine learning and problem solving models known from Computational Intelligence such as Neural Nets, Bayesian networks, Logic Networks, Fuzzy Logic, state machines, evolvable hardware, etc., can be extended to those based on quantum circuits and automata.
  • Computational Intelligence interacting with various aspects of Quantum information theory including error correction, teleportation, encryption/decryption, security, etc.
  • Quantum game theory, applications of quantum games.
  • Using GA, GP and other evolutionary and biological paradigms in all areas of quantum circuits, quantum information and quantum computing.
  • Applications of quantum concepts in Computational Intelligence, Multimedia and Robotics.
  • Quantum entanglement, and applications in communication, computing, information theory, etc.
  • Quantum probability and its applications.

CEC-48 Hybrid Cultural Algorithms: Beyond Classical Cultural Algorithms

Organized by Robert G. Reynolds, Mostafa Ali, Ziad Kobti and Ponnuthurai Nagaratnam Suganthan

Cultural Algorithms are computational models of Cultural Evolution. As such they provide a framework within which experiences of problem solvers embedded in a social fabric influence the collective knowledge of that group, its Culture. Culture is viewed as a network of passive and active knowledge sources. These knowledge sources are able integrate this knowledge, either individually or collectively, into their structure using data mining and machine learning tools. This updated Cultural Knowledge then is used to direct the modifications to individuals and their plans in the population space. Cultural Algorithms are an ideal framework for problems that require large amounts of domain knowledge to direct the collective decisions of individuals in the population. As such Cultural Algorithms have been successfully applied to problems in complex hierarchical systems characterized by large and extensive data sets (big data), many domain constraints, multiple objectives, and multiple agents within a large and spatially distributed social network.

Cultural Algorithm can also provide a flexible framework for hybridization with other socially motivated technologies such as particle swarm optimization, differential evolution, ant colony optimization, and co-evolutionary approaches among others. These hybrid systems have required extension to Classical Cultural Algorithms such as multi-population and multi-belief spaces, and novel approaches to using belief space knowledge to drive evolutionary search. This special session is designed to provide an overview of the diverse hybrid approaches that have been proposed beyond the classical Cultural Algorithm. Cultural Algorithm designers are invited to submit their latest extensions and share a glimpse of the future of Cultural Algorithms.

Scope and Topics

This special session will focus on all aspects of Cultural Algorithms theory and application. Topics of interest may cover, but are not limited to the following:

  • Collective Intelligence
  • Social Intelligence in Networks
  • Bio-informatics applications
  • Multi-Cultural systems and subcultures
  • Multi-Objective Optimization
  • Many Objective Optimization
  • Multi-Agent Systems
  • Ecosystem Modelling and Virtual World Applications
  • Hybrid Systems Learning Systems
  • Distributed Computing
  • Social Intelligence in Games and Auctions
  • Cloud Computing applications
  • Constrained Optimization
  • Real-World Applications
  • Crowd Sourcing
  • Hybrid agent populations: GA, GP, Neural, and Fuzzy agents
  • Education

CEC-49 Genetic Improvement of Software + Search-Based Software Engineering (GI+SBSE)

Organized by Markus Wagner , William Langdon and Brad Alexander

In the past ten years there has been a dramatic increase in work on Search-Based Software Engineering (SBSE), an approach to software engineering in which search-based optimisation algorithms are used to address problems. The approach is attractive because it offers a suite of adaptive automated and semi-automated solutions in situations typified by large complex problem spaces with multiple competing and conflicting objectives. SBSE has been applied to a number of software engineering activities, right across the life-cycle from requirements engineering, project planning and cost estimation through testing, to automated maintenance, service-oriented software engineering, compiler optimisation and quality assessment.

With this special session, we are providing an opportunity to showcase recent breakthroughs in this field.

Scope and Topics

We invite submissions on any aspect of SBSE, including, but not limited to, theoretical results and interesting new applications. The suggested topics cover the entire range of functional and non-functional properties:

  • bandwidth minimisation
  • latency minimisation
  • fitness optimisation
  • energy optimisation
  • software specialisation
  • memory optimisation
  • software transplantation
  • bug fixing
  • multi-objective SE optimisation

CEC-50 Dynamic Multi-objective Optimization

Organized by Mardé Helbig, Kalyanmoy Deb and Andries Engelbrecht

Most real-world optimization problems have more than one objective, with at least two objectives that are in conflict with one another. The conflicting objectives of the optimization problem lead to an optimization problem where a single solution does not exist, as is the case with single-objective optimization problems (SOOPs).Instead of a single solution, a set of optimal trade-off solutions exists, referred to as the Pareto- optimal front (POF) or Pareto front. This kind of optimization problems are referred to as multi-objective optimization problems (MOOPs).

In many real-world situations the environment does not remain static, but is dynamic and changes over time. However, in recent years most research was focussed on either static MOOPs or dynamic SOOPs. When solving dynamic multi-objective optimization problems (DMOOPs) an algorithm has to track the changing POF over time, while finding solutions as close as possible to the true POF and maintaining a diverse set of solutions. Some of the major challenges in the field of dynamic multi- objective optimization (DMOO) are a lack of a standard set of benchmark functions, a lack of standard performance measures, issues with performance measures currently being used for DMOO and a lack of a comprehensive analysis of existing algorithms applied to DMOO.

Therefore, this special session aims to highlight the latest developments in dynamic multi-objective optimization (DMOO) in order to bring together researchers from both academia and industry to address the above mentioned challenges and to explore future research directions for the field of DMOO. We invite authors to submit original and unpublished work on DMOO.

Scope and Topics

Topics of interest include, but are not limited to:

  • DMOO Benchmark functions
  • Performance measures for DMOO
  • Constrained DMOO
  • New DMOO algorithms
  • Comparative studies of DMOO algorithms
  • Theoretical aspects of DMOO algorithms
  • Approaches to handle outlier solutions
  • Real-world applications of DMOO algorithms
It should be noted that we are NOT interested in approaches that change the DMOOP to a DSOOP. These contributions fall within the scope of the special session on dynamic and uncertain environments.

CEC-51 Evolutionary Computation for Human Center Decision Making Systems: Trends and Applications

Organized by Ana Maria Madureira,

Decision Support Systems are interactive software-based system intended to support business and organizational decision-making activities in order to help decision makers to compile information, model business processes, solve problems and make decisions. This special session should address the design of decision support systems with dynamic adaptation and optimization become increasingly important incorporating expert’s knowledge from an Ambient Intelligence perspective. It gives relevance to the idea of human-centered design and the intelligence needed to allow systems to foresee user’s needs and preferences.

This special session intends to present and discuss the integration of recent developments of Evolutionary Computation, Self-Organization, Decision Support Systems, Information System and Human Computer Interaction, Ambient Intelligence, in general.

Scope and Topics

The topics of interest for this special session include, but are not limited to:

  • Evolutionary Computation Optimization-based decision support models
  • Collaborative support systems
  • Human-Computer Interaction for Interactive Systems
  • Web-based Decision Support Systems
  • Intelligent User Interfaces
  • User Modeling preferences
  • Natural interfaces and ambiguity solving techniques
  • Intelligent Agents and Multi-Agent Systems
  • Hybrid Intelligent Systems
  • Self-Organized, Autonomic Computing and Distributed Systems
  • Intelligent Manufacturing Systems
  • Applications: Manufacturing, Logistics and Supply chain management, Biomedical and Bioinformatics, Business, Medicine, Banking, Financing, Social Networks, Workflow

CEC-52 Hidden Complex Networks in Evolutionary Dynamics. Past, Present and Future

Organized by Ivan Zelinka, Guanrong Chen, Ponnuthurai Nagaratnam Suganthan, Andy Adamatzky and René Lozi

Evolutionary computation as well as complex systems dynamics and structure is a vibrant area of research in the last decades. To date, large set of nonlinear complex systems exhibiting chaotic and/or emergent behaviours are observed, analysed and used. They include evolutionary algorithms, as Wright and Agapie proposed in Cyclic and Chaotic Behaviour in Genetic Algorithms in 2001 on GECCO conference. Such algorithms, systems and its mutual fusion form an essential part of science and engineering. Most notable examples include chaos control and synchronization, chaotic dynamics for pseudo-random number generators in evolutionary algorithms, modelling of evolutionary dynamics like complex networks, use of chaos game with evolutionary algorithms and/or use evolution in complex systems design and analysis (evolutions in complex networks). Recently, the study of such phenomena is focused not only on the traditional trends but also on the understanding and analysis of principles, with the new intention of controlling and utilizing it toward real-world applications.

This special session is concerned about evolutionary dynamics as a complex process that can be modelled and analysed by means of complex networks tools. Evolutionary algorithms are complex systems that consist of many interacting units (e.g. individuals, ...) where interactions can be recorded as a virtual complex network. Its attributes can be then analysed, studied and used to better understand dynamical processes inside system under consideration and/or for its control or optimization of its behaviour and structure.

In the session are welcome original research papers discussing new results, based on previous research papers, on PSO, differential evolution, SOMA, GA, ABC and other algorithms whose dynamics was converted into related complex network structure, analysed and used to improve performance of discussed bio-inspired algorithms

Scope and Topics

The aim of this session is to bring together people from fundamental research, experts from various applications of evolutionary algorithms and complex systems, to develop mutual intersections and fusion. Also discussion of possible hybridization amongst them as well as real-life experiences with computer applications will be carried out to define new open problems in this interesting and fast growing field of research. The special session will focus on, but not limited to, the following topics:

  • Evolutionary dynamics as a complex network
  • Analysis of evolutionary dynamics by means of complex network tools
  • Mutual relations amongst evolutionary dynamics, complex network and CML systems
  • Evolutionary dynamics as a feedback loop system, analysis and control
  • Randomness, chaos and fractals in evolutionary dynamics and its impact on algorithm performance.

CEC-53 Multi-Fidelity Design Optimization under Epistemic Uncertainties

Organized by Liqiang Hou, Massimilano Vasile and Edmondo Minisci

Epistemic uncertainties due to lack of knowledge can be found in many real-life design problems. The uncertainties cannot be modeled using the conventional statistical tools, e.g. Gaussian distribution model. Instead, some tools like Evidence Theory or belief function theory can be used to model the uncertainties. Plausibility and belief, which describe the upper and lower bounds of the possible results, are used to evaluate the uncertainty impacts.

In some more complex problems, the epistemic uncertainties takes a more complex form, e.g. the uncertainties with Basic Probability Assignment (BPA) structure of the estimated values and variances. The design optimization under uncertainties can be formulated as multi-objective optimization problem, with the objectives to minimize system function values and the objective to maximize corresponding beliefs. In both cases, a step-like optimal Pareto front should be obtained.

Another problem the designers face frequently in real-life engineering optimization is the fidelity management in the design optimization. A common way to tackle the design problems with multi-fidelity models is to start the optimization with the low fidelity model, and then use the expensive high fidelity model to improve the design solutions. The strategy works well in some problems but have a risk to trap into the pseudo optimal solutions if the values the low level model predicts is not consistent well with the high fidelity ones in that region. Therefore, the model fidelity management strategy for determining when the high fidelity model should be used is required. The strategy can be kriging, space mapping, and trust region method, etc. However, most of the strategies are designed for the single objective design cases, e.g to maximize the lift-drag ratio of the airfoil under specified Mach number. For a more complex design problem with multi-objective optimization under uncertainties, such strategies should be improved before they can be used in the design optimization.

In recent years, methods of optimization under epistemic uncertainties and multi-fidelity models are developed, but in a separate way. However, in many cases, both the epistemic uncertainties and the multi-fidelity models are involved. The session aims to develop efficient and combinatory strategies for these problems, and incorporate efficiently the evidence computation and model fidelity management in the design optimization. As the evidence computation can cost huge amount of computational resources if numbers of epistemic uncertainties are involved, high efficiency approximation techniques are required, and bench mark functions for the design optimization should be developed as well.

Scope and Topics

The design problem under epistemic uncertainties and multi- fidelity models can be found in many practical problems, particularly in aerospace engineering design problems. The issues involved include MOO algorithm, uncertainty modeling, model fidelity management, and the strategies to integrate them to implement the optimization efficiently.

The session seeks to promote the discussion and presentation of novel works related with (but not limited to) the following issues:

  • Application of robust design optimization in engineering problems
  • Uncertainty modeling
  • Parameter reduction technique
  • Model fidelity management
  • Surrogate of expensive model
  • Multi-objective optimization for the problems with step-like optimal front
  • Approximation techniques for evidence computation

CEC-54 Geometrical and Topological methods in Evolutionary Computing

Organized by Vaclav Snasel and Ajith Abraham

One of the biggest challenges in Big Data analysis is to solve the problem of “semantic gaps” between low-level features and high-level semantic concepts. Geometrical and Topological methods are tools for analyzing highly complex data. This methods create a summary or compressed representation of all of the data features to help rapidly uncover critical patterns and relationships in data. The idea of constructing summaries over whole domains of parameter values involves understanding the relationship between geometric objects constructed from data using various parameter values.

The main aim of this Special Session is to explore the new frontiers of big data computing for Geometrical and Topological methods through computational intelligence techniques, in order to more efficiently analyzing highly complex data.

Scope and Topics

The proposed special session aims to bring together theories and applications of Geometrical and Topological methods in Evolutionary Computing to analyze highly complex data. Topics of interest include, but are not limited to:

  • Data sampling and feature subset ranking
  • Feature subset integration
  • Dimensionality reduction for feature summarization
  • Extraction-based summarization
  • Abstraction-based summarization
  • Aided summarization
  • Hybridisation of feature summarization and evolutionary computation
  • Real-world applications of Geometrical and Topological methods in Evolutionary Computing, e.g. images and video sequences/analysis, face recognition, gene analysis, biomarker detection, medical data classification, diagnosis, and analysis, text mining, instrument recognition, power system, sensor systems

CEC-55 Multiobjective Optimization with Surrogate Models

Organized by Bogdan Filipic and Thomas Bartz-Beielstein Carlos A. Coello

Many real-world optimization problems involve multiple, often conflicting objectives and rely on computationally expensive simulations to assess these objectives. Such multiobjective optimization problems can be solved more efficiently if the simulations are partly replaced by accurate surrogate models. Surrogate models, also known as response surface models or meta-models, are data driven models built to simulate the processes or devices that are subject to optimization. They are used when more precise models, such as those based on the finite element method or computational fluid dynamics, spend too much time and resources. While surrogate models allow for fast simulation and assessment of the optimization objectives, they also represent an additional source of impreciseness. In multiobjective optimization, this may constitute a particular challenge when comparing candidate solutions. The aim of this special session is to bring together researchers and practitioners working with surrogate-based multiobjective optimization algorithms to present recent achievements in the field and discuss directions for further work.

Scope and Topics

Prospective authors are invited to submit their original and unpublished work on all aspects of surrogate-assisted multiobjective optimization. The scope of the special session covers, but is not limited to the following topics:

  • State-of-the-art in multiobjective optimization with surrogate models
  • Theoretical aspects of surrogate-assisted multiobjective optimization
  • Novel surrogate-based multiobjective optimization algorithms
  • Comparative studies in multiobjective optimization with surrogates
  • Benchmark problems and performance measures for multiobjective optimization with surrogates
  • Real-world applications of multiobjective optimization using surrogates

FIND US ON

line line

Cross-Disciplinary and Computational


Intelligence Applications Special Sessions


Cross-Disciplinary and Computational Intelligence Applications Special Sessions

CDCI-01 Computational Intelligence for Industry 4.0

Organized by Yun Li, Cindy Goh, Leo Chen and Zhi-Hui Zhan

This Special Session is dedicated to the latest developments of computational intelligence for Industry 4.0, the first a-priori engineered (and the fourth) ‘Industrial Revolution’. Focusing on smart manufacturing and cyber-physical systems so far, efforts in Industry 4.0 have lacked smart design and business elements for manufacture that are necessary in completing this unprecedented upgrade of value chain. Computational intelligence has however provided an extra-numeric, as well as efficiently-numeric, tool to realise this goal. The Special Session therefore encourages and reports applications to Industry 4.0 in the era of interactive cloud computing and data science.

Scope and Topics

Computational intelligence, primarily comprising artificial neural network and learning systems, evolutionary computation, and fuzzy logic and systems, is a set of nature-inspired modelling and optimisation approaches to complex real-world problems, to which traditional approaches such as first principles modelling and statistical or curve fitting are ineffective or incapable of addressing. We are soliciting original research papers or reviews that would shape and advance a smart design and business environment for Industry 4.0. Papers addressing how to revolutionise the way that smart designs are created and smart machines are built, thereby leading to a step improvement in manufacturing autonomy and industrial efficiency, performance and competitiveness, will be most welcome. Main Topics (include but are not limited to):

  • Computer intelligence or machine learning for cyber-physical systems
  • Computer-automated design, machine learning or intelligent search for Industry 4.0
  • Computational intelligence for smart design for smart manufacture
  • Computational intelligence for Industry 4.0 in cloud and big data environments
  • Computational intelligence and data science for marketing in Industry 4.0 value chain
  • Computational intelligence or other learning techniques for Industry 4.0 business informatics
  • Computational intelligence and data science applications to marketing for design
  • Evolutionary distributed or cloud computing for interactive product design and marketing
  • Evolutionary big data interaction for predictive product design and marketing

CDCI-02 Computational Intelligence for Physiological and Affective Computing (CIPAC)

Organized by Faiyaz Doctor, Christian Wagner, Dongrui Wu and Marie-Jeanne Lesot

Affective Computing (AC) is “computing that relates to, arises from, or deliberately influences emotions,” as initially coined by Professor R. Picard (Media Lab, MIT). It has been gaining popularity rapidly in the last decade because it has great potential in the next generation of human-computer interfaces. One goal of affective computing is to design a computer system that responds in a rational and strategic fashion to real-time changes in user affect (e.g., happiness, sadness, etc), cognition (e.g., frustration, boredom, etc.) and motivation, as represented by speech, facial expressions, gestures, physiological signals, neurocognitive performance, etc. Physiological Computing (PC) relates to computation that incorporates physiological signals in order to produce useful outputs (e.g., in computer-human interaction). It mainly differs from AC in the sense that its foremost focus is not the modeling of affect but rather the utilization of physiological information generally.

Practical applications of AC and PC based systems seek to achieve a positive impact on our everyday lives by monitoring, recognising and acting on our emotional states and physiological signals. Integrating these sensing modalities into intelligent and pervasive computing systems will reveal a far richer picture of how our fleeting emotional responses, changing moods, feelings and sensations, such as pain, touch, tastes and smells, are a reaction to or influence how we implicitly or explicitly interact with the environment and increasingly the connected computing artifacts within.

The integration and use of AC and PC raise many new challenges for signal processing, machine learning and computational intelligence. Fuzzy Logic Systems in particular provide a highly promising avenue for addressing some of the fundamental research challenges in AC/PC where most data sources such as: body signals (e.g., heart rate, brain waves, skin conductance and respiration) facial features, speech and human kinematics are very noisy/uncertain and subject-dependent. Clearly however, other key areas of CI research, such as evolutionary learning algorithms and neural network based classifiers provide essential tools to address the significant challenge of AC/PC.

Scope and Topics

The Computational Intelligence and Physiological and Affective Computing special session aims to bring together researchers from the three areas of CI to discuss how CI techniques can be used individually or in combination to help solve challenging AC/PC problems, and conversely, how physiological and affect (emotion) and its modeling can inspire new approaches in CI and its applications. Topics of interest for this special session include but are not limited to:

  • Models of emotion and physiological information
  • Classifiers for physiological information
  • Applications based on/around physiological information
  • Fuzzy set and system based architectures for processing emotions and other affective states
  • Automatic emotion recognition & synthesis from physiological signals, facial expressions, body language, speech, or neurocognitive performance
  • Emotion mining from texts, images, or videos
  • Affective interaction with virtual agents and robots based on fuzzy systems
  • Applications of affective computing in interactive learning, affective gaming, personalized robotics, virtual reality, social networking, smart environments, healthcare and behavioral informatics, etc.

CDCI-03 Computational Intelligence methods for Natural Language Processing

Organized by Keeley Crockett and Joao Paulo Carvalho

Although language, or linguistic expressions, undoubtedly contains fuzziness in nature, very little research has been conducted in related fields in recent years, as it was shown in “A Critical Survey on the use of Fuzzy Sets in Speech and Natural Language Processing”, Proc. of the IEEE WCCI 2012, Brisbane, Australia. This is partly because of the prevalence of probabilistic machine learning technologies in the natural language processing field. However, there has been a growing recognition that fuzziness found in every aspect of human language has to be adequately captured and that recent developments in the fields of computational intelligence such as computing with words can make a contribution. This session will follow on from the successful, special session entitled “Fuzzy Natural Language Processing” which was held at IEEE FUZZ 2015 in Istanbul and IEEE FUZZ 2013 in India and the hybrid special session held at the 2014 IEEE World Congress on Computational Intelligence in Beijing.

The aim of this Special Session is therefore to explore new techniques and applications in the field of computational intelligence approaches to natural language processing.

Scope and Topics

The session will provide a forum to disseminate and discuss recent and significant research efforts in fuzzy, neural and evolutionary methods for natural language processing in addition to hybrid and emerging computational intelligence paradigms. It invites researchers from different related fields and gathers the most recent studies including but not limited to:

  • Fuzzy set models of human language
  • Computational intelligence applications to human language processing
  • Machine learning approaches to human language
  • Computational intelligence approaches to text mining
  • Computational Intelligence simulations of language use
  • Fuzzy ontologies for human language
  • Computational intelligence applications to the semantic web
  • Computing with words within natural language processing
  • Real world computational intelligence inspired natural language processing applications.

CDCI-04 Computational Intelligence in Bioinformatics (CIB)

Organized by Vassilis Plagianakos and Roberto Tagliaferri

Bioinformatics, computational biology, and bioengineering present a number of complex problems with large search spaces. Recent applications of Computational Intelligence (CI) in this area suggest that they are well-suited to this area of research. This special session will highlight applications of CI to a broad range of topics. Particular interest will be directed towards novel applications of CI approaches to problems in these areas. The scope of this special session includes evolutionary computation, neural computation, fuzzy systems, artificial immune systems, swarm intelligence, ant- colony optimization, simulated annealing, and other CI methods or hybridizations between CI approaches. Applications of these CI methods to bioinformatics, computational biology, and bioengineering problems are the main focus of this hybrid special session. There is a clear interest in both the Computational Intelligence community and Biology communities for this special session. This hybrid special session is sponsored by the IEEE CIS BBTC (Computational Intelligence Society - Bioinformatics and Bioengineering Technical Committee).

Scope and Topics

Topics of interest include but are not limited to:

  • Analysis and visualization of large biological data sets
  • biological and medical ontologies
  • biomedical data modelling and mining
  • biomedical model parameterization
  • brain computer interface
  • computational proteomics
  • systems biology
  • ecoinformatics and applications to ecological data analysis
  • emergent properties in complex biological systems
  • gene expression array analysis
  • gene finding
  • genetic networks
  • high-throughput data analysis
  • immuno- and chemo-informatics
  • in-silico optimization of biological systems
  • medical image analysis
  • medical imaging and pattern recognition
  • medicine and health informatics
  • metabolic pathway analysis
  • microarray design or oligonucleotide selection
  • modelling, simulation and optimization of biological systems
  • molecular docking and drug design
  • molecular evolution and phylogenetics
  • molecular sequence alignment and analysis
  • motif and signal detection
  • robustness and evolvability of biological networks
  • single nucleotide
  • polymorphism (SNP) analysis
  • structure prediction and folding
  • systems and synthetic biology
  • treatment optimization

CDCI-05 Computational Intelligence and Games

Organized by Daniel Ashlock and Ruck Thawonmas

Games are an ideal domain to study computational intelligence (CI) methods because they pro- vide affordable, competitive, dynamic, reproducible environments suitable for testing new search algorithms, pattern-based evaluation methods, or learning concepts. They are also interesting to observe, fun to play, and very attractive to students. Additionally, there is great potential for CI methods to improve the design and development of both computer games and non-digital games such as board games. This special session aims at gathering not only leading researchers, but also young researchers as well as practitioners in this field who research applications of computational intelligence methods to computer games.

Scope and Topics

In general, papers are welcome that consider all kinds of applications of CI methods (evolutionary computation, supervised learning, unsupervised learning, fuzzy systems, game-tree search, etc.) to games (card games, board games, mathematical games, action games, strategy games, role-playing games, arcade games, serious games, etc.). Examples include:

  • Adaptation in games
  • Automatic game testing
  • Coevolution in games
  • Comparative studies (e.g. CI versus human-designed players)
  • Dynamic difficulty in games
  • Games as test-beds for CI algorithms
  • Imitating human players
  • Learning to play games
  • Mathematical games such as the snowdrift game or prisoners dilemma
  • Multi-agent and multi-strategy learning
  • Player/opponent modelling
  • Procedural content generation
  • Results of game-based CI competitions
  • Results of open competitions

CDCI-06 Computational Intelligence for Music, Art, and Creativity

Organized by Chuan-Kang Ting, Tatiana Tambouratzis, Francisco Fernández de Vega, Stefanos Kollias, Palle Dahlstedt and Aggelos Pikrakis

Computational intelligence (CI) techniques, including evolutionary computation, neural networks, and fuzzy systems, have gained several promising results and become an important tool in computational creativity, such as in music, visual art, literature, architecture, and industrial design.

Scope and Topics

The aim of this special session is to reflect the most recent advances of CI for Music, Art, and Creativity, with the goal to enhance autonomous creative systems as well as human creativity. This session will allow researchers to share experiences and present their new ways for taking advantage of CI techniques in computational creativity. Topics of interest include, but are not limited to, CI technologies in the following aspects:

  • Generation of music, visual art, literature, architecture, and industrial design
  • Algorithmic design in creative intelligence
  • Application of CI to music theory, analysis, classification/clustering, composition, variation and improvisation
  • Optimization in creativity
  • Development of hardware and software for creative systems
  • Evaluation methodologies
  • Assistance of human creativity
  • Computational aesthetics
  • Emotion response
  • Human-machine creativity

CDCI-07 Advanced Computational Intelligence Methods for Health Technologies and Applications

Organized by Steve S. H. Ling, Xin Xu, H.K. Lam, Hung T. Nguyen, Kit Yan Chan and Rifai Chai

Nowadays, computational intelligence methods play an importance role in the health technology research, which brings together complementary interdisciplinary research practice, in the development of innovative medical devices and biotechnological processes for health applications. In general, feasible results may be obtained by applying traditional artificial intelligence methods to a health application. However, health technologies demand to be more robust, more precise and more efficient. Applying traditional artificial intelligence methods may not achieve multiple goals for a particular health application. Recent research indicates that the advanced computational intelligence methods can help to achieve a more satisfactory performance for a particular health application. With the rapidly growing complexities of health design problems and more demanding quality of health applications, the development of advanced computational intelligence methods for health technologies is hence a critical issue. This special issue intends to bring together researchers to report the latest results or progress in advanced computational intelligence methods for health technologies.

Scope and Topics

The field of interest of this special issue is the application of recent concepts and methods of computational intelligence in health technologies. The topics cover a broad range of health applications, and we are soliciting contributions on (but not limited to) the following aspects:

  • Brain-computer interfaces
  • Bioinformatics
  • Intelligent powered wheelchair
  • Protein-ligand conformation
  • Analysis of heart rate dynamics, cardiovascular disease, diabetes mellitus, neurological disorders
  • Non-invasive instrumentations
  • Early detection of cancer
  • Biomedical signal and image processing, monitoring, and control
  • Physiological modeling
  • ECG/EEG/EMG classification
Advanced computational intelligence methods involved the following technologies but not limited to:
  • Artificial immune systems
  • Deep learning
  • Ensemble learning
  • Evolutionary algorithms
  • Evolutionary programming
  • Reinforcement Learning
  • Fuzzy systems
  • Neural networks
  • Rough sets and random sets
  • Swarm intelligence
  • Support vector machines

CDCI-08 Computational Intelligence Techniques For The Analysis Of Big And Streaming Data In Complex Systems

Organized by Barbara Hammer, D. Frank Hsu and Marios M. Polycarpou

Due to improved sensor technology, increasing storage space, and data availabil- ity, digital data sets are rapidly increasing with respect to size, dimensionality, and complexity. On the one hand, big and streaming data sets are becoming more and more popular in complex systems such as industrial manufacturing processes, surveillance, finance, social networks, or health-care. On the other hand, the dimensionality of data can easily reach a few thousand and data sources are often enriched by auxiliary information which gives crucial clues to avoid overfitting. These facts demand for advanced methods and tools which can cope with these big and complex data with respect to not only its sheer size, but also its often challenging statistical properties such as heterogeneous qual- ity, data trends, presence of rare events, and necessity for strong regularisation.

Scope and Topics This special session will focus on advanced data analysis for big and streaming data which enable a reliable and computationally feasible access to such data sets. Submissions are encouraged according to the following non-exhaustive list of topics:

  • Big data analytics
  • Big data visualization
  • Reliable machine learning for drift and trend
  • Incremenental and lifelong learning
  • Security and privacy in big data
  • Regularization techniques for very high dimensional data
  • Machine learning for heterogeneous and streaming data
  • Constant memory algorithms for data analysis
  • Analysis of sensor networks and social networks
  • Distributed and multiple source machine learning techniques
  • Big data applications e.g. in astronomy, health care, sensor networks
  • Information and data fusion
  • Semi-supervised learning
  • Data correlation vs. information diversity

CDCI-09 Computational Intelligence for Economics and Finance

Organized by Okan Duru and Matthew Butler

Among several topics in computational intelligence, there is a growing interest on economic and financial analysis as well as models for economic and financial management. This special session invites submissions addressing computational advancements and intelligent solutions for economic and financial management in addition to theoretical discussions on the development and use of intelligent techniques.

Scope and Topics

This special session is particularly organized for applications and modelling practice in economics and finance. The scope of this special session includes but not limited to:

  • Financial Management
  • Economic and Financial Decision Making under Uncertainty
  • Artificial Immune Systems
  • Time Series Analysis
  • Forecasting
  • Stock Market Analysis
  • Risk Management
  • Credit Risk Modelling
  • Commodity Markets
  • Pattern Recognition
  • Pricing and Valuation
  • Trading Systems
  • Portfolio Management
  • Algorithmic Trading
  • Data Mining
  • Sentiment Analysis and Behavioral Finance
  • Plasticity of Artificial Systems in Economics and Finance
  • Low frequency / high severity event modelling
  • Big Data
  • Asset Management

CDCI-10 Computational Intelligence Methodologies for Environmental Sustainability and Sustainable Development: Theory and Applications

Organized by Tatiana Tambouratzis, Andreas-Georgios Stafylopatis, Kostas Karatzas and Mikko Kolehmainen

Environmental sustainability has become a topic of particular interest - and concern in - in the last 20 years. Environmental sustainability is focused upon responsible decision-making and action-taking for the protection of the environment, thus boosting the ability of the environment to continue to support life. At the same time, environmental sustainability tackles the issue of developing optimal practices that will reduce - and eventually minimise - the negative impact on the environment. Further to pollution, waste, and energy reduction, environmental sustainability aims at developing processes that will help human societies to become completely sustainable in the future. The use of non-parametric, noise-resistant, and learn-by-example approaches is pertinent to this end, and constitutes the focus of this WCCI’16 Special Session, a sequel to the IJCNN'15 Special Session under the same name.

Scope and Topics

Topics of interest include but are not limited to:

  • Environmental sustainability
  • Sustainable development
  • Decision making
  • Action taking
  • Optimisation
  • Long-, medium- and long-term impact assessment
  • Human activity and the environment
  • Product life-cycle

CDCI-11 Computational Intelligence in Ecological Informatics and Environmental Modelling

Organized by Michael J Watts and Jie Yang

The aim of this special session is to provide a forum for recent research in the application of computational intelligence in the areas of ecological informatics, ecological modelling and environmental modelling. This is a highly topical area and is open to a broad array of methods from the field of computational intelligence, and follows from the successful special session “Applications of Computational Intelligence in Ecological Informatics and Environmental Modelling” held at WCCI 2014.

Ecological informatics and the related field of ecological modelling involve constructing computational models of ecological systems. Environmental modelling is closely related and involves constructing models of the physical environment that biological eco-systems inhabit. Ecological models include such things as the distribution or abundance of particular species, models of the interaction between multiple species, and models of the future development of populations of species. Environmental models cover such topics as the climate and climate change and the detection of landscape features. Models have also been constructed of waste management systems, water quality and drainage systems; Water contamination events; Flood modelling and air pollution.

The amount of data describing global and local environments and the eco-systems that inhabit them is rapidly increasing. As these are highly-complex systems, algorithms from the field of computational intelligence have already been widely applied to modelling this data. Previous work has successfully solved numerous problems in ecological and environmental modeling using artificial neural networks evolutionary algorithms, and fuzzy systems approaches. In each case, computational intelligence methods were shown to be more effective at solving the problem than the alternative methods.

This session is of wide appeal to participants of WCCI 2016 because it involves all three primary fields of interest of the conference: Artificial Neural Networks (ANN), Fuzzy Systems (FS) and Evolutionary Algorithms (EA). It follows on from the successful special session held at WCCI 2014. It is also an emerging area of research as the majority of publications and researchers in this area continue to be ecologists rather than computational intelligence researchers. There is therefore a continuing scope for researchers in computational intelligence to make a strong contribution to this emerging field.

Scope and Topics

Topics relevant to this special session include, but are not limited to, the following applications of computational intelligence, including ANN, FS, and EA:

  • Species distribution and ecological niche modelling
  • Predicting species abundance
  • Remote sensing image analysis and content-based image retrieval for Ecological Informatics and Environmental Modelling
  • Analysis of species assemblages
  • Classification of species
  • Environmental impact assessment
  • Modelling of environmental events including floods
  • Issues in the preparation of ecological data for modelling
  • Modelling of pollutants or contamination in air, land or water
  • Modelling water quality
  • Greenhouse gas emissions modelling and the effects of climate change
  • Modelling the future development of populations
  • Detecting landscape features
  • Modelling water drainage systems
  • Assessment of habitat quality
  • Forecasting of algal blooms
  • Habitat suitability modelling
  • Predicting crop hazards, pests or diseases
  • Modelling interactions between multiple species
  • Identifying landscape features
  • Modelling ecosystem biomass
  • Learning of phenological patterns

CDCI-12 From Big Data to Big Knowledge Using Computational Intelligence in Biomedicine

Organized by Francesco Masulli, Sanaz Mostaghim and Alexandru G Floares

Due to the explosive evolution of Information Technology and Computer Science, Biomedicine entered in Big Data Age, and this is really a scientific revolution, not just a fashion. As always, the technological aspects evolve faster than the scientific community mentality. Transforming Big Data into Big Knowledge and developing a Knowledge-Based Medicine require new visions and approaches. Companies, facing the Big Data challenges, are moving faster in the right direction than the biomedical community, being under a stronger competitive pressure. They were forced to renounce to wishful thinking, like the idea that a few variables, embedded in a few rules, discovered using the old fashion statistics, will give intelligent support for business decisions. We have to do the same for developing adequate diagnosis, prognosis, and response to treatment predictive models/tests.

On the positive side, the biomedical community has to realize that we are already in the Big Knowledge Age too. Curated facts from literature, either manually or by Text/Web Mining, are stored in large repositories and integrated as structured knowledge. Dedicated software tools (e.g., DAVID, Metacore, and Ingenuity Pathways Analysis) allow the users to search for knowledge, which could be represented in biologically meaningful ways, like pathways or networks.

Computational Intelligence (CI) methodologies, tailored to Big Data, and combined with a proper vision of living systems, e.g., as complex dynamical systems or networks of interacting entities, could pave the way to Knowledge-Based Medicine. Precision Medicine should be viewed not only as an increase in measurement's accuracy but also as highly accurate predictive models (Predictive Medicine), discovered from Big Data with CI tools. All the steps of the workflows from Big Data to Big Knowledge could greatly benefit from using all CI methodologies, and this is why this special session is addressed to all of the three sections of the WCCI 2016.

Scope and Topics

All the steps of the workflows from Big Data to Big Knowledge could greatly benefit from using all CI methodologies, and this is why this special session is addressed to all of the three sections of the WCCI 2016. Authors are encouraged to apply CI methods and emphasize how their results could be incorporated into the biomedical domain knowledge. Topics include, but are not limited to:

  • Novel CI approaches for data analysis in biomedicine
  • Scalable computational intelligence tools for biomedicine
  • Deep Learning architecture, representations, unsupervised and supervised algorithms
  • Hardware and Software solutions for Big Data Searching, Storing and Management
  • Microarray and Next-Generation Sequencing Data Preprocessing and Analysis
  • Biomedical imaging preprocessing and Analysis
  • Integration of Microarray NGS OMICS and Clinical Data
  • Structured and Unstructured Data/Text/Web Mining
  • Natural Language Processing in Biomedicine
  • Functional Analysis of differentially expressed and predictive genes
  • Diagnosis, prognosis, and response to treatment predictive models
  • Data Visualization and Visual Analytics
  • Biomedical analysis pipelines, frameworks, and workflows.
  • Pathways and network analysis and visualization
  • Computational drug design and repositioning
  • Clinical Decision Support Systems and Electronic Health Records

CDCI-13 Computational Intelligence in Power System

Organized by Kumarappan N

The demand for electrical energy is growing exponentially and quality and reliability requirements of modern power systems are becoming more and stringent. This special session will focus on the applications of computational intelligence for planning, operation, control, and optimization of electric power systems, in order to provide better secure, stable and reliable system. The computational intelligence include neural computation, evolutionary computation, swarm intelligence, artificial immune systems, ant colony search, pattern recognition, data mining, firefly algorithm, artificial bee colony, etc.

The objective of this special session is to bring together researchers from the academia and industry in the fields of power system engineering and computational intelligence.

Scope and Topics

The special session invites contributions in the areas including, but not limited to, the following:

  • Power system operation
  • Power system control
  • Power system planning
  • Power system analysis
  • Power system stability
  • Power system reliability
  • Power system protection
  • Security assessment
  • Power quality
  • Load frequency control
  • Power sector reforms and restructuring
  • Smart grid

CDCI-14 Computational Intelligence for Security, Surveillance, and Defense

Organized by Derek T. Anderson, Timothy C. Havens and Hussein Abbass

Given the rapidly changing and increasingly complex nature of global security, we continue to witness a remarkable interest within the security and defense communities in novel, adaptive and resilient techniques that can cope with the challenging problems arising in this domain. These challenges are brought forth not only by the overwhelming amount of data reported by a plethora of sensing and tracking modalities, but also by the emergence of innovative classes of decentralized, mass-scale communication protocols and connectivity frameworks such as cloud computing, vehicular networks, and the Internet of Things (IoT). Realizing that traditional techniques have left many important problems unsolved, and in some cases, not addressed, further efforts have to be undertaken in the quest for algorithms and methodologies that can detect and easily adapt to emerging threats.

Scope and Topics

The purpose of this Special Session is to provide a forum for the exchange and discussion of current solutions in Computational Intelligence (e.g., neural networks, fuzzy systems, evolutionary computation, swarm intelligence, and other emerging learning or optimization techniques) as applied to security, surveillance, and defense problems. High-quality technical papers addressing research challenges in these areas are solicited. Papers should present original work validated via analysis, simulation, or experimentation, pertaining, but not limited, to the following topics:
Computational Intelligence for Advanced Architectures for Defense Operations

  • Multi-Sensor Data Fusion
  • Employment of Autonomous Vehicles
  • Intelligence Gathering and Exploitation
  • Mine Detection
  • Situation Assessment
  • Automatic Target Recognition
  • Mission Weapon Pairing and Assignment
  • Sensor Cueing and Tasking
  • Computational Red Teaming
  • Trusted Autonomous Systems

Computational Intelligence for Modeling and Simulation of Defense Operations
  • Logistics Support
  • Mission Planning and Execution
  • Resource Management
  • Course of Action Generation
  • Models for War Games
  • Multi-Agent Based Simulation
  • Strategic Planning
  • Joint Operations
  • Red-Blue M&S

Computational Intelligence for Security Applications
  • Surveillance
  • Human Modeling: Behavior, Emotion, Motion
  • Suspect Behavior Profiling
  • Automated Handling of Dangerous Situations
  • Stationary or Mobile Object Detection, Recognition and Classification
  • Intrusion Detection Systems
  • Cyber-Security
  • Air, Maritime and Land Security
  • Network Security, Biometrics Security and Authentication Technologies
  • Trusted Autonomous Software Systems
  • Spectrum Management

CDCI-15 Computational Intelligence Algorithms for Industrial and Design Engineering Problems

Organized by Mohamed Tawhid and Vimal Savsani

The use of Computational Intelligence techniques like Genetic algorithm, Particle swarm optimization, Differential Evolution, Artificial bee colony optimization, Teaching learning based optimization, Cuckoo search etc., have expand its importance for numerous engineering applications and it is considered as an important tool in the engineering field.

Scope and Topics

The aim of this special session is to reproduce the most current move on of Computational Intelligence techniques for different engineering applications to improve the overall design of the systems. This session is targeted to provide a common platform for the researchers and experts to interact and share their knowledge to take advantage of such techniques in the engineering field. Topics of interest may include, but are not limited to the following highlights:

  • Problems faced in the engineering field
  • Introduction to different Computational techniques
  • Design of algorithms
  • Application of algorithms to different mechanical, electrical, civil and industrial problems
  • Performance assessment of the algorithms
  • Development of effective techniques
  • ulti-objective optimization and its application to different engineering problems.

CDCI-16 Computational Intelligence in Dynamics of ComplexNetworks: Models and Applications

Organized by Chung-Ming Ou and Chung-Ren Ou

Complex networks can be seen everywhere such as biology, chemistry, ecology, economics, physics, and even the Internet. There are many achievements in complex networks based on varied theories and models from mathematics and physics as well. However, the structures and internal properties of complex networks which lead to the major applications in sciences and technologies are still worth exploring. Among them, Computational Intelligence (CI) methodologies can be greatly contributed to solve these issues, in particular, the dynamics of complex networks. Applications of these CI methods to model and simulate complex networks are the main focus of this special session. The scope of the methodologies and ideas may include neural networks, artificial immune systems, swarm intelligence, fuzzy systems and other CI methods or general approaches from applied mathematics, physics, bio-inspired methodologies and control theory.

Scope and Topics

Topics of interest include but are not limited to:

  • analysis and visualization of complex networks
  • artificial immune systems
  • big data and complex networks
  • chaos in complex networks
  • complex adaptive networks
  • complex networks and self-organization
  • control and dynamics of complex networks
  • dynamics of complex networks
  • mathematical modelling and analysis of complex networks
  • modelling and simulation of biological networks
  • robustness and stability of complex networks
  • scale-free complex networks
  • swarm intelligence and complex networks
  • small-world complex network
  • statistical properties of complex networks

CDCI-17 Computational Intelligence Methods Accelerated on Parallel and Distributed Architectures for Applications in Bioinformatics, Computational Biology and Systems Biology

Organized by Paolo Cazzaniga, Daniel Ashlock, Marco S. Nobile and Daniela Besozzi

Research problems in Bioinformatics, Computational Biology and Systems Biology deal with systems at different scales of complexity and granularity (from the inference of single molecular structure to the emergent behavior of genome-wide networks), each one requiring completely different computational methods. Computational intelligence is frequently exploited to devise efficient heuristics solving problems in these disciplines; however, these approaches can be computationally challenging, limiting their applicability to real-world problems. The scope of this special session is to bring together researchers involved in the development of computational intelligence methods applied to Bioinformatics, Computational Biology and Systems Biology, specifically accelerated either by means of conventional architectures (e.g., computer clusters, GRID computing) or by unconventional technologies (e.g., Graphics Processing Units, Many Integrated Core coprocessors, biomimetic devices).

Scope and Topics

The scope of this session includes accelerated Computational Intelligence methods applied to the fields of Bioinformatics, Computational Biology and Systems Biology. Topics of interest include, but are not limited to:

  • analysis and visualization of large datasets
  • analysis and visualization of genome-wide models
  • biomedical model parameterization
  • development of synthetic biological devices
  • drug design
  • emergent properties in complex biological systems
  • flux balance analysis
  • gene expression array analysis
  • high-throughput data analysis
  • medical image analysis
  • medical imaging and pattern recognition
  • metabolic pathway analysis
  • mining of biomedical data
  • modelling, simulation and optimization of biological systems
  • molecular dynamics and molecular docking
  • molecular evolution and phylogenetics
  • molecular sequence alignment and analysis
  • optimization of biological systems
  • prediction and searching of molecular structure and folding
  • spectral analysis

CDCI-18 Computational Intelligence for Unmanned Systems

Organized by Hongwei Mo and Chaoming Luo

An unmanned system(US) is a machine or device that is equipped with necessary data processing units, sensors, automatic control, and communications systems and is capable of performing missions autonomously without human intervention. Unmanned systems include unmanned aircraft, ground robots, underwater explorers, satellites, and other unconventional structures.

Computational Intelligence(CI) includes classical evolutionary computation, neural computation, fuzzy systems, swarm intelligence(Particle Swarm Optimization, Ant Colony Optimization,.etc.) and other new CI methods such as Bee colony optimization algorithms, Biogeography Based Optimization, Firefly algorithms or hybridizations of CI approaches. 

Scope and Topics

This special session aims to cover all subjects of Unmanned Systems relating to the development of automatic machine systems based on CI, which include advanced technologies in unmanned hardware platforms (aerial, ground,underwater and unconventional platforms), unmanned software systems, energy systems, modeling and control, communications systems, computer vision systems, sensing and information processing, navigation and path planning and innovative application case studies.

 Authors are invited to submit their original and unpublished work to this special session. Topics of interest include but are not limited to: 

CI methods solving technical issues underlying the development of unmanned systems.

  • Biologically-inspired computing for US
  • CI for motion planning of unmanned aircraft, ground robots, underwater explorers
  • CI for navigation, mapping and localization of unmanned aircraft, ground robots, underwater explorers
  • CI for image processing of unmanned aircraft, ground robots, underwater explorers
  • Bio-inspired system on US
  • Artificial Neural Networks for US
  • CI on machine learning, intelligent systems design for unmanned hardware platforms (aerial, ground, underwater and unconventional platforms)
  • CI on machine learning and intelligent systems design for unmanned software systems
  • CI for energy systems of US
  • CI for modeling, control and communication systems of US
  • CI for computer vision systems of US
  • CI for sensing and information processing of US
  • Theory and applications of CI and machine learning systems for US
  • Swarm intelligence for unmanned aircraft, ground robots, underwater explorers and other unconventional structures.

CDCI-19 Computational Intelligence in Biometrics and Biometrics Applications

Organized by Yong Xu, Hale Kim, Qinghan Xiao, David Zhang and Fabio Scotti

Biometrics is a technology that focuses on using measurable human physiological or behavioural characteristics to reliably distinguish one person from others. Because of the fuzzy nature of biometrics, there are no two samples that will be perfectly identical. Computational intelligence (CI), primarily based on artificial intelligence, neural networks, fuzzy logic, evolutionary computing, etc., has been exploited to solve biometric problems with promising results. This special session intends to provide an interdisciplinary forum for researchers and practitioners, from industry, government, and academia, to share their research and experiences in the field of computational intelligence in biometrics. The focus of this special session is on innovative, new technologies designed to address important issues in the development of biometrics.

Scope and Topics

Possible topic areas include, but are certainly not limited to the following areas:

  • CI-based biometric solutions for user verification and identification
  • Multi-classifier systems such as multiple biometrics or multi-modal biometrics
  • Data/information fusion for biometric verification
  • Security and privacy issues, biometric cryptography and template protection
  • Spoofing, attacks and countermeasures on biometric systems
  • Adaptive classification systems for biometric recognition
  • Intelligent and evolutionary biometric systems
  • Mobile biometric devices and embedded biometric systems
  • Biometric standards and the development of conformance and interoperability testing tools
  • Human re-identification in a multi-camera surveillance scenario
  • Intelligent sensors and devices for biometric applications

CDCI-20 Computational Intelligence Algorithms in Wireless Sensor Networks

Organized by Anil Kumar, Arun Khosla, Jasbir Singh Saini and Shelly Sachdeva

The Wireless Sensor Networks (WSNs) play a vital role in our society, as they have become the archetype of pervasive technology. WSNs consist of an array of sensors of either the same or diverse types, interconnected by communication network. Fundamental design objectives of the sensor networks include reliability, accuracy, flexibility, cost, effectiveness and ease of deployment. Sensors perform routing function to create single or multi-hop wireless networking to convey data from one to other sensor nodes. The rapid deployment, self-organization and fault-tolerance characteristics of WSNs make them promising for a number of military and civilian applications.

WSN is treated as multi-modal and multi-dimensional optimization problem and addressed through Computational Intelligence involving the minimization of an objective function error.

The main aim of this Special Session is to inquire the new aspects of Computational Intelligence algorithms in WSN to minimize the computational complexity and also improve the performance. 

Scope and Topics

The Special Session aims to bring researchers working in the domain of Computational Intelligence in WSN to an International forum to discuss and explore their latest findings. Topic includes, but are not limited to:

  • Embedding computing platform
  • Data processing and handling
  • Communication Protocols
  • Smart deployment and management
  • Localization
  • Network life time/energy harvesting
  • Time synchronization
  • Security
  • Smart Interactive applications

CDCI-21 Computational Intelligence for Pathology Informatics

Organized by Chao-Hui Huang and Emarene M. Kalaw

Traditionally, detection, diagnosis and prognosis are based on the visual examination of histopathological tissue samples using a microscope and they are labor intensive tasks for pathologists. E.g., The department of pathology in a major hospital (such as Singapore General Hospital) receives about one thousand slides per day. Thus, the emerging technologies of whole slide imaging, digital pathology and virtual microscopy application become more popular as they provide more objective diagnosis, e.g, cancer grading, that will lead to better prognostication. Hence, pathologist’s workload can be reduced.

In addition, the combination of Immunohistochemistry and Virtual Microscopy can be a powerful tool for supporting pathologists’s daily work. For example, the recent research of the oncogenes, PTEN, ERG and C-MYC, are critical in prostate carcinogenesis. However, there are still pitfalls and inter-observer variability in manual assessment on microscope. An automated assessment will be useful for performing diagnosis.

In this session, we will bring pathologists and computer scientists together, discuss the current issues of disease diagnosis in pathology and how the computational intelligence can help in their daily work. We will review the recent development of digital histopathology, present novel methods that will lead to automatic and objective cancer grading in clinical practice and discuss how do these methods of computational intelligence may improve prognostication in the future.

Scope and Topics

The aim of this session is to provide a common platform for the researchers and experts in both fields of pathology and computational intelligence to share their knowledge. Topics of interest may include, but are not limited to:

  • Whole Slide Image Acquisition, Storage and Management
  • Novel Trends of Telepathology
  • 3D Reconstruction of Pathological Images
  • Machine Learning for Pathology Informatics
  • Visual Semantic Learning for Whole Slide Image Analysis
  • Segmentation, Classification and Detection on Whole Slide Images
  • Semantic Mapping between Whole Slide Images and Clinical Reports
  • Computational Intelligence for Cancer Pathology and Genomics
  • Functional Genomics and Immunohistochemistry for Disease Diagnosis in Virtual Microscopy
  • Pathological Imaging Genomics

CDCI-22 Intelligent Control of Unmanned Surface Vehicles

Organized by Meng Joo Er and Ning Wang

Unmanned Surface Vehicles (USVs) are now being deployed in an array of different application areas in the commercial, naval and scientific sectors. For example,, they are currently being used for mine counter-measures, surveying and environmental data gathering. For such vehicles to be capable of undertaking the kinds of mission that are now being contemplated, they require robust, reliable, accurate and adaptable autopilot systems which allow seamless switching between automatic and manual control modes. Such properties in marine control systems are highly required to meet the changes in the dynamic behaviour of the vehicles that may occur owing to the deployment of different payloads, mission requirements and varying environmental conditions.

Modelling and control of USV has been and will be a crucial and challenging issue in both marine engineering sector and control community. Surface vehicles invariably navigate in uncertain environments with unknown disturbances from winds, waves and currents, etc. In this context, mathematical models can at most partially capture the simplified dynamics of USVs since high order hydrodynamic derivatives with respect to essential nonlinearities and external forces can hardly be obtained accurately. Even though nominal models can be derived to certain degree of accuracy, it is still a great challenge to design an effective control law.

Scope and Topics

Previous research works available in literature can be classified into two categories, i.e. model-based and approximation-based strategies. Model-based approaches require system dynamics to be at least partially known so that traditional nonlinear control laws, i.e., feedback linearization, backstepping technique, sliding-mode control (SMC), etc, can be possibly applied. Unfortunately, traditional adaptive control techniques are meaningful only for systems whose nonlinear dynamics and/or uncertainties are linear-in-the-parameters with explicitly defined regressors. Moreover, the surface vehicle dynamics inevitably suffers from complex hydrodynamics, uncertainties and unknown disturbances with respect to external environments, and thereby resulting in great difficulties in using traditional control methods.

In this special session, we will solicit latest research research results and technological know-hows from researchers from all over the world and we hope to have a very robust discussion on the recent developments and futuristic trends pertaining to intelligent control of USVs’. The main topics of this special session include, but are not limited to, the following:

  • Guidance, Navigation and Control (GNC) of surface vehicles
  • Trajectory tracking control strategies of surface vehicles
  • Path following control of surface vehicles
  • Autonomy of surface vehicles cruising
  • Intelligent GNC strategies for surface vehicles under unknown environment
  • Adaptive Robust control of uncertain surface vehicles

CDCI-23 Computational Intelligence for Knowledge and Skills Transfer: Theories, Algorithms and Applications

Organized by Min Jiang, Changle Zhou, Xiangxiang Zeng and Fei Chao

Effective transfer of knowledge would have significant theoretical and practical values. For example, it could enable robots to rapidly gain skills and adapt to new surroundings with relatively low computational cost. However, the path to achieving this goal is obstructed by a number of difficulties related to: computational resource limitations, uncertainties of information acquisition, and omnipresence of ambient noise. This special session aims at the presentations of the latest research activities related to all facets of knowledge and skills transfer, particularly from perspective of computational intelligence and to applications for robots. The special session is open to contributions on any topic directly or indirectly related to computational intelligence in/for knowledge transfer and skill transfer, touching on at least one of the issues mentioned above. Submissions presenting empirical or mathematical results are especially welcomed; but conceptually rigorous and innovative contributions of any kind will be seriously considered, if relevant.

Scope and Topics

Specific topics include, but are not limited to, the following:

  • Theories or Applications of Lifelong Learning, Transfer Learning, Meta-Learning Multitasking
  • Symbolic Knowledge Extraction form Deep Networks
  • Learning from Partial Observations for complex learning environments
  • Multi/Many Objective Optimization
  • Cognitive Robotics, Developmental Robotics, Epigenetic Robotics, Bio-inspired and Cognitive Robotics, and their Applications
  • Theories or Applications of Spiking Neural Models
  • Theory or Application of Structured Learning and Structured Intelligent Systems
  • Knowledge Representation in Human-Level Intelligent System
  • Machine Consciousness; Human-robot Interactions
  • Human-like Intelligent Systems in Manufacturing, Game Playing and Scheduling

CDCI-24 Cognitive Agent and Robotics for Human-Centric Systems: New Models and Challenges

Organized by Giovanni Acampora and Alessandro Di Nuovo

Cognitive Agents are inspired from the human cognitive capabilities to exhibit effective behavior through perception, action, deliberation, communication, and through either individual or social interaction with the environment. These computational agents have the ability to function effectively in circumstances not explicitly planned when the system was designed. This ability makes it specially suited to control complex systems, such as robotics, which are designed to provide services and proactively interact with people and other artificial agents in composite Human-Centric environments. Despite the tremendous potential applications of these “Cognitive Systems” and the subsequent interest from the scientific and industrial communities, several issues and challenges are still open.

The special session aims to attract new developments in the area of cognitive systems, such as novel engineering principles, models and applications of cybernetic systems capable to autonomously improve their capabilities when interacting with the environment in an open-ended process.

As a follow-up, authors of accepted papers will be invited to submit an extended version to the journal IEEE Transaction on Cognitive Systems and Development (new name of Trans. on Autonomous Mental Development.)

Scope and Topics

The special session welcomes all the contributions in the area of Artificial Cognitive Systems applied to human-centric environments, with particular interest (but not limited to) the following topics:

  • Cognitive Systems for human wellbeing
  • Bio-inspired, developmental and Cognitive Robotics
  • Ambient Assisted Living via Cognitive Agents
  • Computational models of human cognition and interaction
  • Engineering applications of Cognitive Systems
  • Artificial Agent cognitive development
  • Natural language understanding
  • Standard tools for the design of Cognitive Systems
  • Medical Systems and Diagnosis
  • Evolutionary Cognitive Robotics

CDCI-25 Computationally Intelligent Methods in Neural Information Processing

Organized by Mufti Mahmud and Amir Hussain

The brain, being the most complex organ in human body, is specialized to process information simultaneously coming from many different sources. The neurons work as basic information processing units in the brain and interconnect to each other to form hierarchical and/or parallel pathways. These pathways are mainly involved in transforming information originated from one or more sources into either action (as in motor movements) or specialized information understood by the brain itself (as in cognitive functions).

To have a detailed and better understanding of these biological phenomena two approaches have been practiced by the research community – experimental and theoretical studies. Also, some theoretical studies are inspired by the nature itself which reframes earlier computational techniques to suggest research on biophysical basis of brain research and its information processing capabilities. Needless to say that most of these studies are results of interdisciplinary research involving medical sciences, life sciences, physical sciences, engineering, and cognitive sciences.

Scope and Topics

The focus of this special session is to address the recent advances in computationally intelligent techniques in processing neural information. Developing intelligent methods capable of deciphering brain’s information processing capability is one the biggest challenges in brain research. The objective of this special session is to provide updated information and a forum for the scientists and researchers who are looking for more relevant information in decoding brain functions using expert and computationally intelligent systems.

This special session is expected to attract papers on recent research progress in the area of intelligent computational methods in processing neural signals. The targeted research topics are, but not limited to, the following:

  • Computationally intelligent tools for analysis of Spikes, LFP, EEG, MEG, MRI/fMRI, PET, and fNIRS;
  • Computationally intelligent methods for modeling and estimating neural signals;
  • Computational intelligence in developing smart BMI and neural prosthesis;
  • Biologically inspired methods for pattern analysis in neuronal signals;
  • Machine learning methods applied to brain research;

CDCI-26 Cognitive Robotics

Organized by Chu Kiong Loo, Janos Botzheim and Naoyuki Kubota

Recently, various types of intelligent robots have been developed for the society of the next generation. In particular, intelligent robots should continue to perform tasks in real environments such as houses, commercial facilities and public facilities. The growing need to automate daily tasks combined with new robot technologies are driving the development of human-friendly robots, i.e., safe and dependable machines, operating in the close vicinity to humans or directly interacting with persons in a wide range of domains. The technology shift from classical industrial robots, which are safely kept away from humans in cages, to robots, which will be used in close collaboration with humans, requires major technological challenges that need to be overcome. A robot should have human-like intelligence and cognitive capabilities to co-exist with people. The study on the intelligence, cognition, and self of robots has a long history. The concepts on adaptation, learning, and cognitive development should be introduced more intensively in the next generation robotics from the theoretical point of view. Fuzzy, neural, and evolutionary computation play important role to realize cognitive development of robots from the methodological point of view. Furthermore, the synthesis of information technology, network technology, and robot technology may bring the brand-new emerging intelligence to robots from the technical point of view. The structurization of information and knowledge is a key topic to support the cognitive development of robots. This special session focuses on the intelligence of robots emerging from the adaptation, learning, and cognitive development through the interaction with people and dynamic environments from the conceptual, theoretical, methodological, and/or technical points of view.

Scope and Topics

The topics of interests in the special session include, but are not limited to:

  • Robot Intelligence
  • Learning, Adaptation, and Evolution in Robotics
  • Human-Robot Interaction
  • Embodied Cognitive Science
  • Perception and Action
  • Intelligent Robots
  • Fuzzy, Neural, and Evolutionary Computation for Robotics - Evolutionary Robotics
  • Soft Computing for Vision and Learning
  • Informationally Structured Space

CDCI-27 Deep Computational Intelligence Models

Organized by Sandeep Paul, Lotika Singh and Apurva Narayan

The computational intelligence (CI) paradigm is a triumvirate of three technologies neural networks, fuzzy logic, and evolutionary algorithms which stresses their seamless integration, resulting in numerous important spin-off commercial applications. The integration of these technologies has assumed various forms such as neuro-genetic, genetic fuzzy and neuro-fuzzy- evolutionary hybrids. In an attempt to address real world complex problems specially with high dimensional data, models with deep architectures are showing promising path for efficient and robust solutions. Deep architecture systems have deep multi-layer structure and represents hierarchical information which is more robust and reusable than the classical neural network approaches.

The hybridization of fuzzy and evolutionary algorithms with such deep architecture systems is in its nascent stage. The inherent challenges like structure and parameter learning, automatic feature extraction and learning algorithms in deep architectures systems are yet to be fully explored. The theoretical analysis of such deep architectures models is another area to be worked on and is much needed.

Scope and Topics

The aim of this special session is to provide a platform to present and deliberate the recent findings and future research directions on the existing deep architecture models, new innovative architectures, various ways of integration of fuzzy and deep neural networks, hybrid approaches to deep architecture systems, evolvable systems, theoretical and empirical analysis and real world applications.

This special session will encourage researchers from academia and industry in exciting and multi-disciplinary field of Deep Computational Intelligence. The topics for the proposed special session include, but are not limited to:

  • Innovative deep architectures/algorithms
  • Hybridization of deep learning models with existing techniques like evolutionary computation, neural networks and fuzzy systems
  • Novel structure and parameter handling techniques for deep architecture models
  • Theoretical and/or empirical results on representation, architecture and learning algorithms in deep networks
  • Parallel implementation of deep learning models
  • Identification of issues involved with real time implementation of deep learning models
  • Advances in hybrid deep learning models for large scale data and real world applications
  • Deep learning in embedded systems/real time critical system

CDCI-28 Applied Computational Intelligence for the Economics of Financial Market Infrastructures

Organized by Biliana Alexandrova-Kabadjova

The development of modern society, the technological innovation and more recently the financial crises have opened up a wide variety of new avenues for economic research, driven primary by the need of authorities for specific policy oriented studies.

The economic analysis in Financial Market Infrastructures is one of this new policy-oriented fields. Financial Market Infrastructures (FMIs) are economic platforms built with the purpose to facilitate the clearing, settlement, and recording of monetary and other financial transactions. The prime elements that now a day form part of the FMIs are payment systems (PS), central securities depositories (CSD), securities settlement systems (SSS), central counterparties (CCP), and trade repositories (TRs). The Principles for Financial Market Infrastructures1 (the Principles), are the pillars of operational rules with the aim of guaranteeing the stability of financial markets and increasing the efficiency of these economic platforms.

Aftermath of the financial crisis we have witnessed multifaceted regulatory efforts around the world. In order to guaranty stability, much attention has been paid to additional bank’s requirements (at the micro-level) and the so-called macro-prudential approach. However, in order to form a proper systemic analysis of the financial industry the level of financial market infrastructures (FMIs) deserves attention in its own right. It was referred by Dr. Martin Diehl as the backbone of the financial system (Diehl, 2016). FMIs are in charge of clearing, settlement and recording of transactions. Studying the economics of FMIs differs from both the micro- and the macro-level as on one side those infrastructures are transactional systems with many participants having direct access to them and forming rich network of interactions. On the other side FMIs are single institutions that are connected through the overlapping set of participants having access to them and through the funds of transfers that flow from one FMI to another. FMIs are themselves complex systems whose operations cannot be modelled analytically, but in order to be studied properly, researchers have to applied complex models. FMIs serve different financial institutes and very often are of systemic importance to the whole financial system, as they determine the efficiency, stability and overall the reliability of financial industry. This field is becoming one of the key channels for conducting macro-prudential policy. Nevertheless very few specialists worldwide are doing research and defining the agenda for the future.

Under this scenario, a field that has proven to be inherently cross-country and intra-disciplinary, the aim of the special session is to promote research on the applications of computational intelligence and studies aimed at sharing insights on best practices for analyzing FMIs.

Scope and Topics

In order to implement stress testing, monitoring and early warning indicators, testing new policy and operational rules topics of interest include, but are not limited to:

  • evolutionary computation
  • network topology
  • agent-based modeling
  • transactional analysis
  • simulations
  • nature-inspired computational models
  • meta-heuristic techniques

CDCI-29 Computational Intelligence in Marketing and Social Sciences

Organized by Raymond Chiong Yukun Bao Manuel Chica and Sergio Damas

Computational intelligence has a long history of applications to marketing and plays an important role in establishing the interdisciplinary pool of methodologies employed in marketing science research. For example, evolutionary algorithms, artificial neural networks, support vector machines and fuzzy logic have been used in demand forecasting, direct marketing and cross selling, among others. Expert systems have been used for decision support in brand management, and data mining has become a core component of customer relationship management in marketing. Likewise, the use of computational intelligence in social science research allows heightened understanding of the dynamics of complex systems. Agent-based modelling, using agents whose intelligence includes full-blown creativity thanks to their ability to learn and to adapt, is revealing information about such systems that has never before been supported.

The purpose of this special session is to bring together the computational intelligence community as well as researchers from marketing and social sciences to set up visions on how state-of-art computational intelligence techniques can be and are used for insightful marketing and social science analysis, and how marketing and social scientists can contribute in promoting new applications with computational intelligence.

Scope and Topics

We invite submissions of original, previously unpublished papers with topics on, but not limited to, the following:

Technical issues include (but not limited to)

  • Evolutionary Algorithms
  • Artificial Neural Networks
  • Support Vector Machines/Support Vector Regression
  • Fuzzy Logic
  • Expert Systems
  • Data Mining
  • Knowledge Discovery
  • Business Intelligence
  • Machine Learning
  • Agent Based Modelling
Issues of marketing and social sciences include (but not limited to)
  • Sales/Demand Forecasting
  • Response Modelling
  • Retailing and Pricing
  • Advertising
  • Customer Relationship Management
  • Brand Management
  • Social Marketing
  • Cognitive and Behavioural Sciences
  • Computational Social Science
  • Politics, Public Policy and Law

CDCI-30 Human Symbiotic Systems

Organized by Tomohiro Yoshikawa and Yoichiro Maeda

This special session aims at discussing the basic principles and methods of designing intelligent interaction with the bidirectional communication based on the effective collaboration and symbiosis between the human and the artifact, i.e. robots, agents, computer and so on.

We aims at encouraging the academic and industrial discussion about the research on Human-Agent Interaction (HAI), Human-Robot Interaction (HRI), and Human-Computer Interaction (HCI) concerning Symbiotic Systems. Reflecting the fact that this society covers a wide range of topics, in this session we invite the related researchers from a variety of fields including intelligent robotics, human-machine interface, Kansei engineering and so on.

Scope and Topics

Topics of interest include, but are not limited to theory and application of:

  • Human-Agent Interaction (HAI)
  • Human-Robot Interaction (HRI)
  • Human-Computer Interaction (HCI)
  • Social Communication or Interaction
  • Partner or Communication Robots
  • Hospitality Robots
  • Human Interface Systems
  • Cooperative Intelligence
  • Kansei Engineering

CDCI-31 Computational Intelligence, Nature-Inspired Learning and Big Data

Organized by Asim Roy, Plamen Angelov, Marley Vellasco, Adel Alimi, G. Kumar Venayagamoorthy, Juyang Weng, Leonid Perlovsky and De-Shuang Huang

The aim of this special session is to promote new advances and research directions in efficient and innovative algorithmic approaches to analyzing big data (e.g. deep learning, nature-inspired and computational intelligence approaches), implementations on different computing platforms (e.g. neuromorphic, GPUs, clouds, clusters) and applications of big data to solve real-world problems (e.g. weather prediction, transportation, energy management).

Scope and Topics

Topics of interest include, but are not limited to theory and application of:

  • Autonomous, online, incremental learning – theory, algorithms and applications in big data
  • High dimensional data, feature selection, feature transformation – theory, algorithms and applications for big data
  • Scalable algorithms for big data
  • Learning algorithms for high-velocity streaming data
  • Deep learning algorithms
  • Machine vision and big data
  • Brain-machine interfaces and big data
  • Cognitive modeling and big data
  • Embodied robotics and big data
  • Fuzzy systems and big data
  • Evolutionary systems and big data
  • Evolving systems and big data
  • Neuromorphic hardware for scalable machine learning
  • Parallel and distributed computing for big data (cloud, HPC, GPUs, clusters, etc.)
  • Big data and healthcare/medical applications
  • Big data and energy systems/smart grids
  • Big data and transportation systems
  • Big data in large sensor networks
  • Big data and machine learning in computational biology, bioinformatics
  • Big data and cloud computing, large scale stream processing on the cloud

CDCI-32 Robust Data Mining Techniques through Hybrid Metaheuristic (RDMM)

Organized by Simon Fong, Xin-she Yang, Thomas Hanne and Sabah Mohammed

Hybridizing techniques embrace the advantages of more than any one of them alone. On one hand metaheuristic approaches are general strategies for guiding heuristic procedures usually for improving the efficiency of optimization methods. Recently there have been strong momentums centered on metaheuristics research in computer science communities especially those of evolutionary computing, swarm intelligence which taps on the power of collective and bio-inspired collaborative behaviours for distributed search. Many contemporary algorithms and their applications to solve computationally intensive problems have emerged, ranging from swarm intelligence methods inspired by bee pollination to wolf-pack hunting.

While novel metaheuristics are being developed from time to time, hybrid versions of them across the data mining techniques are not uncommon. Hybridization comes in two major directions – data mining approaches are combined within an optimization process, and vice-versa. In the first case, data mining is the core function with the objective of analyzing data and revealing the hidden patterns, just as if it is in its original form. The data mining function is wrapped by the iterative process of optimization, driven by some metaheuristics for the sake of stochastically finding the optimal data mining result out of many possible runs. On the other hand, data mining methods are used as a part of the heuristic search in the metaheuristics. In this case, search patterns and search directives are learnt from data mining the past trials during the optimization process. So the heuristic searches are better guided by incorporating the knowledge learnt from the heuristic trails; hence it enhances the optimization results at the end.

The main focus of this special session is to investigate new methods and desirable properties of the new hybrids resulted from combining metaheuristics and data mining, either in metaheuristics wrapping data mining or data mining enhancing the metaheuristic searches. Most metaheuristics strategies have already been applied to data mining tasks but there are still open research lines to improve their usefulness.

Scope and Topics

This special session is intended to serve as a platform for exchanging the latest progresses along these two types of hybridization of the two important computational techniques. Sharing of experiences of applications using hybrid metaheuristics and data mining are encouraged too, both from academia and industries in the following theoretical and application areas (but not limited):

Algorithms and metaheuristics

  • Ant colony optimization
  • Artificial immune systems
  • Bee algorithms
  • Cuckoo search
  • Differential evolution
  • Genetic algorithms
  • Genetic programming
  • Firefly algorithm
  • Harmony search
  • Particle swarm optimization
  • Simulated annealing
  • Tabu search
  • and others
Applications based on Nature-Inspired Computing and Metaheuristics
  • Decision support systems
  • Swarm intelligence and optimization
  • Data mining
  • Scheduling optimization
  • Big data analytics
  • Intelligent information technology
  • Intelligent agents and nature-inspired computing
  • Real-world applications

FIND US ON

line line

General Information

The IEEE WCCI 2016 will offer a number of tutorials aimed at researchers, students and practicing professionals. All tutorials will be held on 24 July 2016. Traditionally, tutorials attract a broad range of audiences, including professionals, researchers from academia, students, and practitioners, who wish to enhance their knowledge in the selected tutorial topic. Tutorials offer a unique opportunity to disseminate in-depth information on specific topics in computational intelligence.


IEEE WCCI 2016 Tutorial Schedule

IEEE WCCI 2016 Tutorials will be held at the Vancouver Convention Centre (West Building) on Sunday, 24 July 2016. It is free of charge for registered participants of the Congress. The tutorial schedule is given below:

Location Room 202 Room 203 Room 204 Room 205 Room 206 Room 207 Room 208-9 Room 215-6
8:30 - 9:00 am Coffee Break
9:00 - 11:00 am IJCNN-3 IJCNN -4 IJCNN -5 FUZZ -3 FUZZ -4 FUZZ -5 CEC -1 CEC -2
Learning from Unstructured Data Streams in Cybersecurity Learning in Non-Stationary Environments Physics of the Mind Type-2 Fuzzy Ontology and Fuzzy Markup Language for Real-World Applications A Sum-of-Squares Framework for Fuzzy Systems Modelling and Control:Beyond Linear Matrix Inequalities Type-2 Fuzzy Sets and Systems Meta-heuristics for Multi-objective Optimization Applying Evolutionary Computation in Industrial Practice
11:00 - 12:30 pm Lunch Break (On Your Own)
12:30 - 2:30 pm IJCNN-6 IJCNN -8 IJCNN -7 CEC -3 CEC -4 CEC -5 FUZZ -2 FUZZ -7
Passive and Active Control for a Lower-limb Rehablitation Robot Multiview Learning Patent Citation Network Analysis Automatic Algorithm Configuration: Methods, Applications and Perspectives Advances in Particle Swarm Optimization Search Based Software Engineering: Foundations, Recent Advances, Challenges and Future Research Directions Fuzzy Logic and Machine Learning: A Tutorial Dynamic Fuzzy Neural Networks: Architectures, Algorithms and Applications
2:30 - 3:00 pm Coffee Break
3:00 - 5:00 pm CEC -6 CEC -8 CEC -7 FUZZ -1a FUZZ -6a IJCNN -9 IJCNN -2 IJCNN -1
Differential Evolution with Ensembles and Topologies Theory of Evolutionary Algorithms Niching Methods for Multimodal Optimization Big Data: Technologies and Computational Intelligence Approaches Computer Vision: A Computational Intelligence Perspective Application of Graphs and Network Theory for the Interpretation of Brain Imaging Data Spiking Neural Networks: The Machine Learning Approach Data Stram Mining
5:00 - 6:30 pm FUZZ -1b FUZZ -6b
Big Data: Technologies and Computational Intelligence Approaches Computer Vision: A Computational Intelligence Perspective

Due to free tutorials, number of seats for each tutorial will be limited to its capacity. They will be given at the first come first take basis. Please arrive early to assure a seat.


IJCNN 2016 Tutorials

  • IJCNN-01 Data Stream Mining
  • IJCNN-02 Spiking Neural Networks: The Machine Learning Approach
  • IJCNN-03 Learning from Unstructured Data Streams in Cybersecurity
  • IJCNN-04 Learning in Non-Stationary Environments
  • IJCNN-05 Physics of the Mind
  • IJCNN-06 Passive and Active Control for a Lower-limb Rehabilitation Robot
  • IJCNN-07 Patent Citation Network Analysis
  • IJCNN-08 Multiview Learning
  • IJCNN-09 Application of Graphs and Network Theory for the Interpretation of Brain Imaging Data

FUZZ-IEEE 2016 Tutorials

  • FUZZ-IEEE-01 Big Data: Technologies and Computational Intelligence Approaches
  • FUZZ-IEEE-02 Fuzzy Logic and Machine Learning: A Tutorial
  • FUZZ-IEEE-03 Type-2 Fuzzy Ontology and Fuzzy Markup Language for Real-World Applications
  • FUZZ-IEEE-04 A Sum-of-Squares Framework for Fuzzy Systems Modeling and Control: Beyond Linear Matrix Inequalities
  • FUZZ-IEEE-05 Tutorial on Type-2 Fuzzy Sets and Systems
  • FUZZ-IEEE-06 Computer Vision: A Computational Intelligence Perspective
  • FUZZ-IEEE-07 Dynamic Fuzzy Neural Networks: Architectures, Algorithms and Applications

IEEE CEC 2016 Tutorials

  • CEC-01 Meta-heuristics for Multi-objective Optimization
  • CEC-02 Applying Evolutionary Computation in Industrial Practice
  • CEC-03 Automatic Algorithm Configuration: Methods, Applications, and Perspectives
  • CEC-04 Advances in Particle Swarm Optimization
  • CEC-05 Search Based Software Engineering: Foundations, Recent Advances, Challenges and Future Research Directions
  • CEC-06 Differential Evolution with Ensembles and Topologies
  • CEC-07 Niching Methods for Multimodal Optimization
  • CEC-08 Theory of Evolutionary Algorithms

IJCNN 2016 Tutorials



Data Stream Mining

Organized by Leszek Rutkowski

In recent years data stream mining became a very challenging and widely studied issue in computer science community. This topic is being developed as a response to the substantial growth of data amount which needs to be processed and analyzed in many fields of human activity. Among many others this includes for example traffic control, network monitoring, fraud detection in bank transactions and issues associated with image recognition like robotic vision or object tracking.

Data streams are potentially infinite sequences of data elements, which arrive to the system often with very high rates. Therefore the standard data mining algorithms used for static data are not directly applicable in this field. They require significant modifications beforehand or totally new dedicated algorithms have to be developed. In this tutorial we will review existing data stream mining algorithms and present new results recently published in [1-4]. The tutorial will consist of 10 parts:

  1. In this part the basic concepts concerning stream data mining are presented. Data mining tasks are described and the properties of data streams are discussed. A survey on state-of-the-art of data stream classification methods (including Hoeffding Trees, VFDT and CVFDT algorithms, ADWIN and FISH algorithms DDM and EDDM drift detectors, ensembles approaches) is provided.
  2. This part focuses on the decision tree induction algorithms ad their applications to data stream mining. The static decision tree algorithm is described in details. Then the Hoeffding trees and the VFDT algorithm are described. It is explained why the application of Hoeffding's inequality is inadequate for information gain (e.g. ID.3 or C4.5 algorithms) and Gini gain (the CART algorithm).
  3. In this part a new statistical tool for establishing the splitting criteria is presented, i.e. the McDiarmid's inequality. Based on the McDiarmid's inequality two splitting criteria (information gain and Gini gain) for decision tree nodes are established.
  4. This part introduces a new impurity measure, i.e. the misclassification error, never proposed for data stream mining in the literature so far. The property of misclassification error is presented and it is compared with information entropy and Gini index. Then a new statistical tool based on the Gaussian approximation is presented for establishing splitting criteria for decision tree nodes.
  5. In this part a new hybrid splitting criterion is proposed. It merges together the splitting criterion for Gini index based on the McDiarmid's inequality and the criterion for misclassification-based split measure based on the Gaussian approximation. It is shown that the resulting decision tree provides satisfactory classification results and reveals the advantages of both component of splitting criteria.
  6. In this part a method, based on the properties of the normal distribution, to choose the best attribute to make a split in considered node, is presented. The scheme of the part is similar to that in part 3.
  7. In this part various ensemble algorithms for stream data mining are presented.
  8. In this part new methods for choosing optimal number of ensemble components will be described.
  9. In this part all the proposed methods are theoretically compared and the results of experimental analysis are depicted.
  10. This part summarizes the tutorial and briefly outlines the ideas for future work.

Biography

Leszek Rutkowski received the MSc and PhD degrees from the Technical University of Wroclaw, Poland, in 1977 and 1980, respectively. Since 1980, he has been with the Technical University of Czestochowa, where he is currently a professor and director of the Institute of Computational Intelligence. From 1987 to 1990, he held a visiting position at the School of Electrical and Computer Engineering, Oklahoma State University. His research interests include data stream mining, big data analysis, neural networks, fuzzy systems, computational intelligence, pattern classification, and expert systems. He has published more than 200 technical papers, including 22 in various series of IEEE Transactions. He is the author of the following books: Computational Intelligence (Springer, 2008), New Soft Computing Techniques for System Modeling, Pattern Classification and Image Processing (Springer, 2004), Flexible Neuro-Fuzzy Systems (Kluwer Academic, 2004), Methods and Techniques of Artificial Intelligence (2005, in Polish), Adaptive Filters and Adaptive Signal Processing (1994, in Polish), and coauthor of two others (1997 and 2000, in Polish) Neural Networks, Genetic Algorithms and Fuzzy Systems and Neural Networks for Image Compression. He is the president and founder of the Polish Neural Networks Society. He was an associate editor of the IEEE Transactions on Neural Networks (1998-2005) and IEEE Systems Journal (2007-2010). He is an editor-in-chief of the Journal of Artificial Intelligence and Soft Computing Research, and he is on the editorial board of the IEEE Transactions on Cybernetics, International Journal of Neural Systems, International Journal of Applied Mathematics and Computer Science and International Journal of Biometric. He is a recipient of the IEEE Transactions on Neural Networks 2005 Outstanding Paper Award. He served in the IEEE Computational Intelligence Society as the chair of the Distinguished Lecturer Program (2008-2009) and the chair of the Standards Committee (2006-2007). He is the founding chair of the Polish chapter of the IEEE Computational Intelligence Society, which won the 2008 Outstanding Chapter Award. In 2004, he was elected as a member of the Polish Academy of Sciences. In 2004, he was awarded by the IEEE Fellow membership grade for contributions to neurocomputing and flexible fuzzy systems.

Spiking Neural Networks: The Machine Learning Approach

Organized by Nathan Scott and Nikola Kasabov

This tutorial introduces spiking neural networks (SNN), their methods, implementations and applications. SNNs use principles of information processing, also characteristic of the brain. Information is represented in the form of many sequences of spatio-temporal potentials (spikes) that are transferred between many neurons through connections. When applied for data modelling SNNs have the potential of compact representation of space and time, fast information processing, time-based and frequency-based information representation, efficient learning and generalisation on complex data, predictive spiking activity that can trigger in advance a necessary response. SNNs can revolutionise computing in general and that is why SNN have been chosen as the main information processing paradigm for the development of new computing, neuromorphic systems in the EU Human Brain Project, the USA Brain project, and others. The tutorial will include materials and demonstrations organized in three parts:

  1. Introduction to SNN: Methods of data encoding and learning in SNN
  2. SNN model design and system implementation:
    All steps of the design of a machine learning model for complex temporal or spatio-/spectro temporal data modelling are discussed and illustrated including:
    • Input data transformation into spike sequences;
    • Learning spatio-temporal spike sequences in a scalable 3D SNN reservoir;
    • On-going learning and classification of data over time;
    • Dynamic parameter optimization;
    • Predictive data modelling with the SNN, so that once a SNN is trained on whole input patterns related to a given outcome (occurrence of an event), the SNN predicts this event earlier in time and accurately when new, partial data is entered;
    • Adaptation of the SNN model on new data, possibly in an on-line/ real time mode;
    • SNN model visualisation and interpretation for a better understanding of the data and the processes that generated it. Implementations of the SNN models as both software and neuromorphic hardware systems are discussed and demonstrated.
  3. SNN applications for temporal-, or spatio-/spectro-temporal data modeling and pattern recognition:
    Applications across domain areas are demonstrated, including:
    • moving object recognition;
    • predictive modelling systems;
    • brain data modeling;
    • neuromorphic hardware;

Biography

Nikola Kasabov

Professor Nikola Kasabov is Fellow of IEEE, Fellow of the Royal Society of New Zealand and DVF of the Royal Academy of Engineering, UK. He is the Director of the Knowledge Engineering and Discovery Research Institute (KEDRI), Auckland. He holds a Chair of Knowledge Engineering at the School of Computing and Mathematical Sciences at Auckland University of Technology. Kasabov is a Past President and Governor Board member of the International Neural Network Society (INNS) and also of the Asia Pacific Neural Network Assembly (APNNA). He is a member of several technical committees of IEEE Computational Intelligence Society and a Distinguished Lecturer of the IEEE CIS (2012-2014) He is a Co-Editor-in-Chief of the Springer journal Evolving Systems and has served as Associate Editor of Neural Networks, IEEE TrNN, IEEE TrFS, Information Science, J. Theoretical and Computational Nanosciences, Applied Soft Computing and other journals. Kasabov holds MSc and PhD from the TU Sofia, Bulgaria. His main research interests are in the areas of neural networks, intelligent information systems, soft computing, bioinformatics, neuroinformatics. He has published more than 600 publications that include 15 books, 180 journal papers, 80 book chapters, 28 patents and numerous conference papers. He has extensive academic experience at various academic and research organisations in Europe and Asia, including: TU Sofia, University of Essex, University of Otago, Advisor-Pofessor at the Shanghai Jiao Tong University, Guest Professor at ETH/University of Zurich. Prof. Kasabov has received the APNNA ‘Outstanding Achievements Award’, the INNS Gabor Award for ‘Outstanding contributions to engineering applications of neural networks’, the EU Marie Curie Fellowship, the Bayer Science Innovation Award, the APNNA Excellent Service Award, the RSNZ Science and Technology Medal, and others. He has supervised to completion 38 PhD students.

Nathan Scott

Nathan Scott is a Postdoctoral Fellow KEDRI, researching the theory and practice of SNN and neuromorphic an systems. He holds a PhD in Computer Science, Bachelor of Computer and Information Sciences (First Class Honours), BCIS (Software Development) and BBus degrees from Auckland University of Technology. Nathan is an AUT Vice Chancellor's Scholar, recipient of Top Graduate awards, the Dean's highest achievement award and of other study awards. He is a member of the IEEE CIS and SigProc Societies, and a member of the IEEE CIS Neural Networks Task Force on Education. He has given a number of invited talks internationally, including tutorials at IJCNN and ICONIP conferences, and IEEE Summer Schools, and chaired a number of conference Special Sessions on SNN. He currently teaches undergraduate courses in computer graphics and embedded computing.

Learning from Unstructured Data Streams in Cybersecurity

Organized by Seiichi Ozawa

Increasing the maliciousness and the diversity of cyber-attacks is one of the most concerned issues in recent years. There are various kinds of cyber-threads such as malware infection, DDoS attacks, probing to find security vulnerability, phishing, and spam mails to lure malicious web site, which intend to steal money/important information and to stop/disturb public services, etc. All these attacks are conducted via communication on the Internet, in which unstructured information is delivered or broadcasted in the form of packet data. Therefore, to utilize machine learning in detection, classification, and prediction of cyber-attacks, we need to consider how unstructured data streams should be formulated as structured data that are fit for machine learning schemes. Because such unstructured data often have no class label, no information on useful features, and no available training set in advance, which pose us big hurdles in applying machine learning methods. In many cases, a solution to the above issues is problem-dependent. However, there might be some rules of thumb in the process of learning from unstructured data streams efficiently and effectively. I hope to share my experiences on the above topics with the audience.

This tutorial includes the following topics in cybersecurity:

  1. How to define a supervised or unsupervised learning problem from a target task,
  2. How to define a feature set for unstructured data,
  3. How to collect data streams and put the labels automatically (i.e., autonomous mechanism to obtain training data),
  4. How and when should a learning system interact with human experts?,
  5. How to utilize visualization techniques for learning unstructured data.
I hope that my talk would give some hints to the persons who are working and want to work in future for practical problems dealing with unstructured data streams.

Biography

Seiichi Ozawa received the B.E. and M.E. degrees in instrumentation engineering from Kobe University in 1987 and 1989, respectively. In 1998, he received his Ph.D. degree in computer science from Kobe University. He is currently a professor with the Department of Electrical and Electronic Engineering, Graduate School of Engineering, Kobe University, Kobe, Japan. From March 2005 to February 2006, he was a visiting researcher at Arizona State University. His current research interests are machine learning, incremental learning, online feature extraction, pattern recognition, and cyber security. He has published more than 124 journal and refereed conference papers, and 8 book chapters/monographs. He is currently an associate editor of IEEE Trans. on Neural Networks and Learning Systems, Evolving Systems, and Pattern Analysis and Applications Journal. He is also serving as a vice-president of Japan Neural Network Society (JNNS) and board members of Asia Pacific Neural Network Assembly (APNNA) and the Institute of Systems, Control and Information Engineers (ISCIE). He is a member of neural networks technical committee of IEEE Computational Intelligence Society. He is currently working as a General Co-Chair of ICONIP2016 in Kyoto, Japan, and served as the special session chairs of ICONIP2013 and WCCI2014, and also served as the technical committee members of many international conferences.

Learning in Non-Stationary Environments

Organized by Cesare Alippi and Manuel Roveri

Many real-world machine learning applications assume the stationarity hypothesis for the process generating the data. This assumption guarantees that the model learnt during the initial training phase remains valid over time and that its performance is in line with our expectations. Unfortunately, this assumption does not truly hold in the real world, in many cases representing only a simplistic approximation.

Current research in machine learning aims at removing/weakening the stationary assumption so that time variance is detected as soon as possible and suitable actions activated afterwards. In this direction, the literature addressing the learning in nonstationary environments classifies existing approaches as passive or active depending on the learning mechanism adopted to deal with the process evolution. Passive approaches rely on a continuous adaptation of the application without explicitly knowing whether a change has occurred or not, while, in active approaches, triggering mechanisms, e.g., Change Detection Tests (CDTs) or Change Point Methods (CPMs), are considered to detect a change in the process generating the data. Once the change has been detected the application might require (self) adaptation to track the system evolution.

The tutorial will introduce and contrast passive and active approaches by providing those details the scholar and the practitioner need to be able to design machine learning applications working in nonstationary environments.

Biography

Cesare Alippi

Cesare Alippi received the degree in electronic engineering cum laude in 1990 and the PhD in 1995 from Politecnico di Milano, Italy. Currently, he is a Full Professor of information processing systems with the Politecnico di Milano. He has been a visiting researcher at UCL (UK), MIT (USA), ESPCI (F), CASIA (RC), USI (CH).
Alippi is an IEEE Fellow, Vice-President education of the IEEE Computational Intelligence Society (CIS), Board of Governors member of the International Neural Networks Society, Associate editor (AE) of the IEEE Computational Intelligence Magazine, past AE of the IEEE-Tran. Neural Networks (2005-2012), IEEE-Trans Instrumentation and Measurements (2003-09) and member and chair of other IEEE committees. He was awarded the 2016 IEEE CIS Outstanding Transaction on Neural networks and Learning Systems award, the 2013 IBM Faculty award; the 2004 IEEE Instrumentation and Measurement Society Young Engineer Award; in 2011 has been awarded Knight of the Order of Merit of the Italian Republic. Current research activity addresses adaptation and learning in non-stationary environments and Intelligence for embedded systems.

Manuel Roveri

Manuel Roveri received the Dr.Eng. degree in Computer Science Engineering from the Politecnico di Milano (Milano, Italy) in June 2003, the MS in Computer Science from the University of Illinois at Chicago (Chicago, Illinois, U.S.A.) in December 2003 and the Ph.D. degree in Computer Engineering from Politecnico di Milano (Milano, Italy) in May 2007. Currently, he is an associate professor at the Department of Electronics and Information of the Politecnico di Milano. He has been visiting researcher at Imperial College London (UK). Manuel Roveri is an Associate Editor of the IEEE Transactions on Neural Networks and Learning Systems and served as chair and member in many IEEE subcommittees. He received the 2016 IEEE CIS Outstanding Transaction on Neural networks and Learning Systems award. He is the co-organizer of the IEEE Symposium on Intelligent Embedded Systems in 2014 and organizer and co-organizer of workshops and special sessions at IEEE-sponsored conferences. Current research activity addresses adaptation and learning in non-stationary environments and intelligence for embedded systems and cognitive fault diagnosis.

Physics of the Mind

Organized by Leonid I. Perlovsky

The presentation focuses on mathematical models of the fundamental principles of the mind-brain neural mechanisms and practical applications in several fields. Physics of the mind is an extension of neural networks towards more realistic modeling of the mind from perception to the entire mental hierarchy including higher cognition and emotion. Big data and autonomous learning algorithms are discussed for cybersecurity, gene-phenotype associations, medical applications to disease diagnostics, financial predictions, data mining in distributed data bases, learning of patterns under noise, interaction of language and cognition in mental hierarchy. Mathematical models of the mind-brain are discussed for mechanisms of concepts, emotions, instincts, behavior, language, cognition, intuitions, conscious and unconscious, abilities for symbols, functions of the beautiful and musical emotions in cognition and evolution. This research won National and International Awards.

A mathematical and cognitive breakthrough, dynamic logic is described. It models cognitive processes “from vague and unconscious to crisp and conscious,” from vague representations, plans, thoughts to crisp ones. It resulted in more than 100 times improvements in several engineering applications; brain imaging experiments at Harvard Medical School, and several labs around the world proved it to be a valid model for the brain-mind processes. New cognitive and mathematical principles are discussed, language-cognition interaction, function of music in cognition, and evolution of cultures. How does language interact with cognition? Do we think using language or is language just a label for completed thoughts? Why the music ability has evolved from animal cries to Bach and Lady Gaga? The presentation briefly reviews past mathematical difficulties of computational intelligence and new mathematical techniques of dynamic logic and neural networks implementing it, which overcome past limitations.

The presentation discusses cognitive functions of emotions. Why human cognition needs emotions of beautiful, music, sublime. Dynamic logic is related to knowledge instinct and language instinct; why are they different? How languages affect evolution of cultures. Language networks are scale-free and small-world, what does this tell us about cultural values? What are the biases of English, Spanish, French, German, Arabic, Chinese; what is the role of language in cultural differences?

Relations between cognition, language, and music, are discussed. Mathematical models of the mind and cultures bear on contemporary world, and may be used to improve mutual understanding among peoples around the globe and reduce tensions among cultures.

Biography

Dr. Leonid Perlovsky is Professor of Psychology Northeastern University, CEO LPIT, past Visiting Scholar at Harvard University School of Engineering and Applied Science, Harvard University Medical School, Principal Research Physicist and Technical Advisor at the Air Force Research Laboratory (AFRL). He led research projects on neural networks, modeling the mind and cognitive algorithms for integration of sensor data with knowledge, multi-sensor systems, recognition, fusion, languages, music cognition, and cultures. As Chief Scientist at Nichols Research, a $0.5B high-tech organization, he led the corporate research in intelligent systems and neural networks. He served as professor at Novosibirsk University and New York University; as a principal in commercial startups developing tools for biotechnology, text understanding, and financial predictions. His company predicted the market crash following 9/11 a week before the event. He is invited as a keynote plenary speaker and tutorial lecturer worldwide, including most prestigious venues, like Nobel Forum, published more than 500 papers, 17 book chapters, and 5 books, including “Neural Networks and Intellect,” Oxford University Press, 2001 (currently in the 3rd printing) and “Cognitive Emotional Algorithms” Springer 2011. Dr. Perlovsky participates in organizing conferences on Neural Networks, CI, Past chair of IEEE Boston CI Chapter; serves on the Editorial Board for ten journals, including Editor-in-Chief for “Physics of Life Reviews”, IF=9.5, T-R rank #4 in the world, on the INNS Board of Governors, a past chair of the INNS Award Committee. He received National and International awards including the Gabor Award, the top engineering award from the INNS; and the John McLucas Award, the highest US Air Force Award for basic research.

Passive and Active Control for a Lower-limb Rehabilitation Robot

Organized by Zeng-Guang Hou

Stroke, traumatic brain injury (TBI), and spinal cord injury (SCI) are most important reasons that cause nervous system damage, and thus lead to physical disabilities. Because of long process of nervous system recovery, the patients would suffer permanent disability without effective treatment and rehabilitation. At present, physical therapy, occupational therapy and exercise therapy are most popular clinical treatments for rehabilitation, and they have been proven helpful to the recovery of patients’ nervous system and limb functions. However, the vast majority of rehabilitation hospitals still carry out the above treatments manually or using simple rehabilitation medical devices. For traditional exercise therapy, most hospitals still use simple cycling devices for passive training which is very limited because of single training mode and fixed training trajectory of such machines. Since the training process for patients of neurological damage is repetitive, it is expected to improve the current status of rehabilitation by using robotics, and also it would accelerate the rehabilitation process for patients and reduce therapists’ labor intensity. We will mainly address the system design of a reclining type rehabilitation robot for lower limbs, and also studied the passive training, active training and assistance training control methods for the needs of neurological rehabilitation and motor function of lower limbs for SCI or stroke patients.

Biography

Dr. Hou received the B.E. and M.E. degrees in electrical engineering from Yanshan University (formerly Northeast Heavy Machinery Institute), Qinhuangdao, China, in 1991 and 1993, respectively, and the Ph.D. degree in electrical engineering from Beijing Institute of Technology, China, in 1997. From July 1999 to May 2004, he was an Associate Professor with the State Key Laboratory of Management and Control of Complex Systems, Institute of Automation, Chinese Academy of Sciences, where he has been a full Professor since June 2004, and the Deputy Director of the Laboratory since 2006. From September 2003 to October 2004, he was a Visiting Professor at the Intelligent Systems Research Laboratory, College of Engineering, University of Saskatchewan, Saskatoon, SK, Canada.
His current research interests include neural networks, robotics for rehabilitation and minimally invasive surgery, and intelligent control systems. He has published over 100 papers in referred journals and conference proceedings. He has over 20 patents. He is the recipient of the YangJiaChi Award by the Chinese Automation Society in 2010, Distinguished Graduate Student Supervisor Award by Chinese Academy of Sciences in 2010, and Excellence Youth Funds by the Natural Science Foundation of China in 2012.
Dr. Hou currently serves as an Associate Editor of IEEE Transactions on Cybernetics, Neural Networks, Acta Automatica Sinica, and Control Theory and Applications. He served IEEE WCCI as the Publicity Chair in Vancouver in 2006, the Publication Chair Hong Kong in 2008, and the Local Arrangement Chair in Beijing in 2014.

Patent Citation Network Analysis

Organized by Péter Érdi

The network of patents connected by citations is an evolving graph, which represents the innovation process. A patent citing another implies that the cited patent reflects a piece of previously existing knowledge that the citing patent builds upon. It will be explained why ad how to use specific algorithms to extract relevant information about the patent citation network. The tutorial consists of three parts:

  1. Topology and evolution of patent citation networks

    The understanding of the development of the patent citation network contributes to discover the rules that govern its growth. The topology and development of patent citation network can be explained with a generic model of network development that combines preferential attachment and ageing. Microsocopic level studies helped to measure the ''attractivness'' of a patent, as the function of its age and the number of citation already has obtained.

  2. Prediction of emerging field of technologies

    At mesoscopic level the analysis has been extended to subclasses, and it was demonstrated by adopting clustering algorithms that it is possible to detect and predict emerging new technology clusters. Analysis of the patent citation network based on its hierarchical structure can be used:
    • to provide a general predictive analytic methodology, which is able to identify structural changes in the patent cluster system and reveal precursors of emerging new technological fields
    • to test and validate the predictive force of the new methodology based on historical examples of new class formation
    • to identify specific mechanisms of the recombination process and formation of new classes
    • to scan the database to identify "hot spots" that may reflect incipient development of new technological clusters
    • to help deciding how to allocate resources.

  3. Recursive ranking: from web pages to patents

    In the theory of social networks centrality measures were constructed to rank network nodes based on their topological importance. The majority of these centrality measures reflect that either there is a connection between a pair of nodes or there isn't, and thus the elements of the adjacency matrix of the graph are zeros or ones. Another family of measures is related to the spectral properties of the adjacency matrix taking into account the importance of the neighbors and led to the construction of the celebrated recursive PageRank algorithm

    The spirit of the PageRank algorithm can be extended to patent citation analysis as well. A patent is useful if it contains significant, reusable information. The proof of the usefulness is that the patent is being cited by other patents. The importance of a patent is measured by the frequency of the citation of the given patent by other patents; however, the weight of these citations is not equal: more important are those citations that are cited by important citations By adopting a citation-based recursive ranking method for patents the evolution of new fields of technology can be traced.. The theory and application of rankng algorithms will be discussed in context of the analogy og biological and technological evolution.

Biography

Péter Érdi serves as the Henry R Luce Professor of Complex Systems Studies at Kalamazoo College. He also has a position of a research prfoessor at Wigner Research Centre for Physics, Hungarian Academy of Sciences in Budapest. In addition, he is the founding director of the BSCS (Budapest Semester in Cognitive Science), a study abroad program. He is a member of the Board of Governors of the International Neural Network Society, some other committees, and serves now also as the Editor-in-Chief of the Cognitive Systems Research. PE's main scientific field is computaional neuroscience, but he is also active on the field of computational social science and other areas of complex systems research.

Multiview Learning

Organized by Roberto Tagliaferri

Multi-view learning is concerned with the problem of machine learning from data represented by multiple distinct feature sets. The recent emergence of this learning mechanism is largely motivated by the property of data from real applications where examples are described by different feature sets or different views, for example: Bioinformatics (microarray gene expression, RNASeq, PPI, gene ontology, etc.); Neuroinformatics (fMRI, DTI); Internet of Things; Web Mining.

In 2013 Sun proposed a Multi View Learning Taxonomy in which there were several main issues: Dimensionality Reduction; Semi Supervised learning; Supervised learning; Clustering; Active Learning; Ensemble Learning, Transfer Learning. Orthogonal to this approach one can analyze the Multi-View learning paradigm with respect to the steps when data is integrated: Early Integration; Intermediate Integration; Late Integration. Each view is analyzed on its own and the results are then fused together.

The tutorial is divided into two parts:

  • An overview of the literature with a presentation of the most significant models/approaches
  • A presentation of applications of Multi-View approach to two of the most interesting research area: Bioinformatics and Neuro-imaging. The aim of the tutorial is to give the audience not only an introduction to Multi-View Learning approaches but also to show some useful applications to real world complex data sets.

Biography

From 1986 to 2000 Roberto Tagliaferri was researcher of Computer Science and Cybernetics, from March 2000 to October 2006 was Associate Professor of Computer Science and since November 1 2006 has been full Professor at the Faculty of Sciences, University of Salerno.

He co-organized Italian and international workshops on Neural Networks and Bioinformatics from 1988. He was co-editor of the proceedings of WIRN from 1995 to 2005. He has been co-editor of special issues on international journals: Neural Networks (2003), International Journal of Approximate Reasoning (2008), Artificial Intelligence in Medicine (2009), International Journal of Knowledge Engineering and Soft Data Paradigms (2010), BMC Bioinformatics (2015). He is Associate Editor of the IEEE Transactions on Cybernetics and of Source code in Medicine and Biology of Biomed Central. He is senior member of the IEEE "Computational Intelligence" and "System, Man and Cybernetics" societies and of INNS. He is chair of the Italian Chapter of the IEEE CIS. He is author of more than 180 scientific publications, includ9ing more than 60 papers on International Journals, co-editor of 17 Proceedings, H-index 23 and i10-index 40 on Google Scholar.

Application of Graphs and Network Theory for the Interpretation of Brain Imaging Data

Organized by Robert Kozma

This self-contained tutorial reviews the intensively developing field of network science and graph theory for researchers working in the field of neural networks and computational intelligence, who are interested in gaining insight into recent progress in the field. We introduce theoretical foundations and computational modeling tools to interpret large-scale brain imaging data. Neural and cognitive networks in the brain are viewed as large-scale graphs, which evolve in time in response to the dynamically changing environmental conditions. Results of the interpretation of brain imaging data are used to design more intelligent computational devices and autonomous robots.

Graph theoretical approaches have been extremely useful in the past 20 years to describe structural and functional properties of large-scale networks, including the world-­wide-web, social networks, ecological networks, biological systems, etc. Our focus here is on neural systems, in particular on brain networks. Functional connections between cortical areas are described using various brain imaging tools, including fMRI, MEG, EEG, and ECOG. These results show breakthroughs in our understanding of brain networks and dynamics, and produce details of the connectome.

Neural correlates of higher cognitive functions are described as the result of brain imaging experiments, which reveal mechanisms of intelligent behavior and the occurrence of moments of deep insight, the "aha" moment. Graph theoretical tools are employed to interpret experimental findings with brain imaging, and to contribute to better understanding of normal and abnormal brain conditions. Specific application areas include intelligent brain-computer interfaces with biofeedback modalities to relieve stress conditions and to enhance relaxation, as well as help the disabled and the elderly.

This tutorial addresses the following topics:

  1. Graph theory modeling tools:
    • Random graph theory, Erdos-Renyi, Strogatz-Watts, Albert-Barabasi models,
    • Scale-free structure and dynamics, black swans and dragon kings,
    • Percolation and characterization of phase transitions.
  2. Review of experimental results on brain imaging:
    • Single cell experiments and electrocorticogram population probes,
    • Noninvasive techniques, fMRI, MEG, scalp EEG,
    • Description of the connectome.
  3. Computational approaches to large-scale brain networks:
    • Lattice models, hierarchy of models,
    • Neuropercolation approach,
    • Implementation of massive computational models.
  4. Critical behavior in neural and cognitive networks:
    • Phase transitions in cognitive processing,
    • Interpretation of phase transitions as “aha” moments,
    • Neural correlates of higher cognition and consciousness.
  5. Applications in biological and engineering areas:
    • Advanced Brain Computer Interface techniques,
    • Brain connection diseases, Alzheimer, Parkinson, TBI, epilepsy,
    • Support quality of life for elderly.

Biography

Robert Kozma (Fellow of IEEE, Fellow of INNS) is Professor of Mathematical Sciences and Director of the Center of Large-Scale Integration and Optimization Networks (CLION), the University of Memphis, TN, USA. Research in his CLION is focused on developing advanced optimization techniques based on biologically motivated and cognitive principles of large-scale networks. He has published 8 books, 300+ papers, has 3 patent disclosures. His research has been supported by NSF, NASA, JPL, AFRL, DARPA, FedEx, and by other agencies. He is President-Elect (2016) of the International Neural Network Society, and serves on the Board of Governors of IEEE Systems, Man, and Cybernetics (SMC, 2016-2018). Previously, he has served on the AdCom of the IEEE Computational Intelligence Society (2009-2012) and the Board of Governors of the International Neural Network Society (2007-2012). He has been General Chair of IJCNN2009, Atlanta, USA. He is Associate Editor of Neural Networks, Neurocomputing, Cognitive Systems Research, and Cognitive Neurodynamics. Dr. Kozma is the recipient of the INNS “Gabor Award” (2011); the “Alumni Association Distinguished Research Achievement Award” (2010), and he has been a “National Research Council (NRC) Senior Fellow” (2006-2008).

Dr. Kozma holds a Ph.D. in Physics (Delft, The Netherlands, 1992), two M.Sc. degrees (Mathematics, Budapest, Hungary, 1988; Power Engineering, Moscow, Russia, 1982). He worked as Research Fellow at the Hungarian Academy of Sciences, Budapest, Hungary. He has been on the faculty of Tohoku University (Japan), Otago University (New Zealand), and had a joint appointment with the Division of Neurobiology and the EECS at UC Berkeley, and has held visiting positions at NASA/JPL, Sarnoff Co., Princeton, NJ; Lawrence Berkeley Laboratory (LBL); and AFRL, Dayton, OH.

FIND US ON

line line

FUZZ-IEEE 2016 Tutorials


Big Data: Technologies and Computational Intelligence Approaches

Organized by Isaac Triguero and Francisco Herrera

Abstract

In the era of the information technology, the problem of managing big data applications is becoming the main focus of attention in a wide variety of disciplines such as science, business, industry, etc, because of enormous increment of data generation and storage that has taken place in the last years. Analyzing and extracting knowledge from such volumes data becomes a very interesting and challenging task for most of the standard computational intelligence techniques that may not be properly adapted to the new space and time requirements. Thus, we must consider new paradigms to develop scalable algorithms.

The paradigm MapReduce introduced by Google allows us to carry out the processing of large amounts of information. Its open source implementation, named Hadoop, led the development of a popular platform with a wide use. Recently, new frameworks as Apache Spark are emerging. Different machine learning libraries are developed for these frameworks, such as Mahout (Hadoop) and MLlib (Spark).

In this tutorial we will provide a gentle introduction to the problem of big data, including a formal definition and the issues that it brings to the society today, as well as the presentation of recent technologies (Hadoop ecosystem, Spark). Then, we will dive into the field of big data analytics, explaining the challenges that come to computational intelligence techniques and introducing machine learning libraries such as Mahout and MLlib. Afterwards, we will go across two of the main topics of the WCCI 2016: fuzzy modeling and evolutionary models in the big data context. Some big data cases of study will be presented for evolutionary feature selection/weighting and fuzzy rule learning, including the associated software.

Table of contents:

  • A gentle introduction to big data. Technologies and Applications
  • Big Data Analytics
  • Fuzzy Modeling for Big Data
  • Evolutionary Models for Big Data Analytics
Click here to get more information, papers and software related to the tutorial.

Biography

Issac Triguero

Isaac Triguero received the M.Sc. and Ph.D. degree in Computer Science from the University of Granada, Granada, Spain, in 2009 and 2014, respectively. He is currently post-doctoral researcher at the Inflammation Research Center of the Ghent University, Ghent, Belgium. His research interests include data mining, data reduction, biometrics, evolutionary algorithms, semi-supervised learning, bioinformatics and big data learning.

Francisco Herrera

Francisco Herrera is a Professor in the Department of Computer Science and Artificial Intelligence at the University of Granada, Spain. He has been the supervisor of 38 Ph.D. students. He has published more than 300 journal papers (H-index 101) that have received more than 36000 citations (Scholar Google). He is co-author of the books "Genetic Fuzzy Systems" (World Scientific, 2001) and "Data Preprocessing in Data Mining" (Springer, 2015).
He currently acts as Editor in Chief of the international journals "Information Fusion" (Elsevier) and “Progress in Artificial Intelligence (Springer). He acts as editorial board member of a dozen of journals, among others: International Journal of Computational Intelligence Systems, IEEE Transactions on Fuzzy Systems, IEEE Transactions on Cybernetics, Information Sciences, Knowledge and Information Systems, Fuzzy Sets and Systems, Applied Intelligence, Knowledge-Based Systems, Memetic Computing, and Swarm and Evolutionary Computation.
He is a Fellow of the European Coordinating Committee for Artificial Intelligence and the International Fuzzy Systems Association. He has been given many awards and honors for his personal work or for his publications in journals and conferences. His areas of interest include, among others, data science, data preprocessing, cloud computing and big data.

Fuzzy Logic and Machine Learning: A Tutorial

Organized by Hamid Tizhoosh

Abstract

In this tutorial we will talk about the state of the art of fuzzy algorithm in machine learning. In the first part, we will review fuzzy algorithms when they are applied on typical machine-learning taks such as search, classification, approximation and learning. In the second part, the relationship between fuzzy methods and other machine-learning approaches are reviewed whereas hybrid schemes will be in foreground. In both parts, relevant literature will be reviewed. Matlab examples will be executed to display the effect of major methods for relevant applications such as data mining, signal processing, image analysis, and big data. Links to online resources will be included in the material which also contains the source codes and the presentation slides.

Structure of the tutorial:

  • Brief History of Fuzzy Logic
  • Brief History of Machine Learning
  • Fuzzy Algorithms for Search, Classification, Approximation and Learning
  • Fuzzy Algorithms and Other Machine-Learning Methods
  • Applications: Data Mining, signal processing, image analysis, and big data
  • Matlab Hands-on Code Development

Biography

Dr. Hamid Tizhoosh received the MSc degree in electrical engineering with a major in computer science from University of Technology, Aachen, Germany , in 1995. From 1993 to 1996, he worked at Management of Intelligent Technologies Ltd., Aachen. Germany in the field of industrial image processing. Dr. Tizhoosh received his PhD degree from University of Magdeburg, Germany, in 2000 with the subject of fuzzy processing of medical images. Dr. Tizhoosh was active as the scientist in the engineering department of IPS(Image Processing Systems Inc., now Photon Dynamics), Markham, Canada, until 2001. For six months, he visited the Knowledge/Intelligence Systems Laboratory, University of Toroanto, Canada.
Since September 2001, Dr. Tizhoosh is a faculty member at the Deparment of Systems Design Engineering, University of Waterloo, Canada. At the same time, he has been the Chief Technology and Chief Executive Officer of Segasist Technologies, a software company (Toronto, Canada) developing innovative software for medical image analysis. His research encompasses machine learning, fuzzy logic and computer vision. Dr. Tizhoosh has extensive experience in medical imaging including portal (megavoltage) imageing, x-rays, MRI and ultrasound. He has been a member of the European Union Projects INFOCUS and ARROW for radiation therapy to improve the integration of online images within the treatment planning of cancer patients. Dr. Tizhoosh has extensively published on fuzzy techniques in image processing. He is the author of two books, 14 book chapters, and more than 100 journal/conference papers.

Type-2 Fuzzy Ontology and Fuzzy Markup Language for Real-World Applications

Organized by Chang-Shing Lee Giovanni Acampora and Yuandong Tian

Abstract

Many different real-world applications with a high-level of uncertainty proved the good performance of the type-2 fuzzy sets (T2 FSs). In this tutorial, we will present three read-world applications, including, game-playing, dietary assessment, and IRT-based e-learning based on type-2 fuzzy ontology and fuzzy markup language. Below is their brief descriptions:

  • The game of Go is a long-history board game that is much more complex than chess. The uncertainties of this game will be higher when the board size gets bigger. Therefore, a sample of games played against a computer is used to estimate the human’s strength. In order to increase the precision, the strength of the computer is adapted from one move to the next by increasing or decreasing the computational power based on the current situation and the result of games. T2FSs with parameters optimized by a genetic algorithm is used to estimate the rank in a stable manner, independently of the board size. T2FS-based adaptive linguistic assessment system infers the human performance and presents the results using the linguistic description.
  • Balanced diet means the intake of each necessary nutrient meets its adequate demand and actual caloric intake balances with calories burned. Additionally, making a diversity of choice from various types of food is also essential to reduce the risk of developing various chronic diseases. Different people have a different goal and it is hard to measure how healthy the eaten meal is for those who are not the domain experts on the diet. The linguistic knowledge discovery mechanism presents the discovered linguistic meaning about the meal’s health level to show the involved subjects how to make a personalized diet linguistic recommendation. This type of information about the eaten meal can provide the subjects with a reference to gradually improve their unhealthy eating habit and then become healthier and healthier.
  • Owing to advanced technical progress in information and communication technology, computerized adaptive assessment becomes more and more important for the personalized learning achievement. Additionally, there are many students learning their academic studies via on-line education platform with many learning materials, for example, KWS Learn (https://sites.google.com/site/kwslearn/); however, how to select learning materials that exactly fit to their competence is not easy for them. According to the response data from the conventional test and three-parameter logistic (3PL) model of the item response theory (IRT), we will combine IRT with fuzzy markup language (FML) for an adaptive assessment application and will propose an intelligent adaptive assessment platform (IAAP) to allow students to do adaptive testing to assess their learning ability. After learning, students provide a feedback to the IAAP and starts next learning iteration to achieve the goal of students’ learning progress.

Biography

Chang-Shing Lee

Chang-Shing Lee (SM’09) received the Ph.D. degree in Computer Science and Information Engineering from the National Cheng Kung University, Tainan, Taiwan, in 1998.
He is currently a Professor with the Department of Computer Science and Information Engineering, National University of Tainan, where he is the Dean of Research and Development Office from 2011 to 2014. His current research interests include adaptive assessment, intelligent agent, ontology applications, Capability Maturity Model Integration (CMMI), fuzzy theory and applications, and machine learning. He also holds several patents on Fuzzy Markup Language (FML), ontology engineering, document classification, image filtering, and healthcare.
He was the Emergent Technologies Technical Committee (ETTC) Chair of the IEEE Computational Intelligence Society (CIS) from 2009 to 2010 and the ETTC Vice-Chair of the IEEE CIS in 2008. He is also an Associate Editor or Editor Board Member of International Journals, such as IEEE Transactions on Computational Intelligence and AI in Games (IEEE TCIAIG), Applied Intelligence, Soft Computing, Journal of Ambient Intelligence & Humanized Computing (AIHC), International Journal of Fuzzy Systems (IJFS), Journal of Information Science and Engineering (JISE), and Journal of Advanced Computational Intelligence and Intelligent Informatics (JACIII). He also guest edited IEEE TCIAIG, Applied Intelligence, Journal of Internet Technology (JIT), and IJFS.
Prof. Lee was awarded the outstanding achievement in Information and Computer Education & Taiwan Academic Network (TANet) by Ministry of Education of Taiwan in 2009 and the excellent or good researcher by National University of Tainan from 2010 to 2013. Additionally, he also served the general chair of The 2015 Conference on Technologies and Applications of Artificial Intelligence (TAAI 2015), general co-chair of 2015 IEEE Conference on Computational Intelligence and Games (IEEE CIG 2015), the program chair of the 2011 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE 2011), and the competition chair of the FUZZ-IEEE 2013. He is also a member of the Program Committees of more than 50 conferences. He is a senior member of the IEEE CIS, a member of the Taiwanese Association for Artificial Intelligence (TAAI), and the Software Engineering Association Taiwan. He was a member of the standing committee of TAAI from 2011 to 2014 and one of the standing supervisors of Academia-Industry Consortium for Southern Taiwan Science Park from 2012 to 2013.

Giovanni Acampora

Dr. Giovanni Acampora (Senior Member, IEEE) received the Laurea (cum laude) and Ph.D. degrees in computer science from the University of Salerno, Salerno, Italy, in 2003 and 2007, respectively. Currently, he is a Reader in Computational Intelligence at the School of Science and Technology, Nottingham Trent University, Nottingham, U.K. From July 2011 to August 2012, he was in a Hoofddocent Tenure Track in Process Intelligence at the School of Industrial Engineering, Information Systems, Eindhoven University of Technology (TU/e), Eindhoven, The Netherlands. His research interests include: fuzzy logic and applications, evolutionary computation, ambient intelligence, forensic intelligence, reputation and trustiness in e-commerce, and so on. In this context, he designed and developed the Fuzzy Markup Language, an XML-based environment for modeling transparent fuzzy systems and he is chairing the IEEE 1855 WG, the working group devoted to make FML as the first IEEE Standard in the area of computational intelligence. Dr. Acampora served as chair and vice-chair of the IEEE CIS Standards Committee. Dr. Acampora acts as Area Chair of the IEEE International Conference on Fuzzy Systems, Chair of the Workshop on Computational Intelligence Tools at IEEE Symposium Series on Computational Intelligence 2015. He acted as Program Committee Member of several conferences in the area of robotics and artificial and computational intelligence. He is acting as General co-Chair for Fuzz-IEEE 2017.

Yuandong Tian

Yuandong Tian is a Research Scientist in Facebook AI Research, working on Deep Learning and Computer Vision. Prior to that, he was a Software Engineer in Google Self-driving Car team in 2013-2014. He received Ph.D in Robotics Institute, Carnegie Mellon University on 2013, Bachelor and Master degree of Computer Science in Shanghai Jiao Tong University. He is the recipient of 2013 ICCV Marr Prize Honorable Mentions for his work on global optimal solution to nonconvex optimization in image alignment.

A Sum-of-Squares Framework for Fuzzy Systems Modeling and Control: Beyond Linear Matrix Inequalities

Organized by Kazuo Tanaka

Abstract

This talk presents a comprehensive treatment of system-theoretical approaches to fuzzy systems modeling and control. These approaches are enabled by a progression of design frameworks from the well-known convex linear matrix inequality (LMI) based design to non-convex sum-of-squares (SOS) based synthesis. Today, there exists a large body of literature on fuzzy model-based control using LMIs. A key feature of LMI-based approaches is that they result in simple, natural and effective design procedures as alternatives or supplements to other nonlinear control techniques that require special and rather involved knowledge. The LMI-based design approaches entail obtaining numerical solutions by convex optimization methods such as the interior point method. Though LMI-based approaches have enjoyed great success and popularity, there still exist a large number of design problems that either cannot be represented in terms of LMIs, or the results obtained through LMIs are sometimes conservative. A post-LMI framework is the SOS-based approaches for control of nonlinear systems using polynomial fuzzy systems and controllers, which includes the well-known Takagi-Sugeno fuzzy systems and controllers as special cases. The SOS framework has been extensively applied to guaranteed-cost control, observer design, discrete system stabilization, etc. To obtain a polynomial fuzzy controller by solving design conditions efficiently, non-convex design conditions are transformed to convex design conditions. However the transformation often results in some challenging issues in SOS-based approaches. Conversely, non-convex design conditions can avoid the transformation problems, but it is difficult to solve non-convex design conditions efficiently. To this end, this talk presents a most recent result on an efficient numerical technique to deal with non-convex design conditions.

The research covered in this talk has been conducted in our laboratory at the University of Electro-Communications (UEC), Tokyo, Japan, in collaboration with Prof. Hua O. Wang and his laboratory at Boston University, Boston, USA. Throughout the talk, it will be reflected upon how to bridge enabling fuzzy model-based control frameworks with system-theoretical approaches in the development of toolkits for control of nonlinear systems.

Biography

Professor Kazuo Tanaka is currently a Professor in Department of Mechanical Engineering and Intelligent Systems at the University of Electro-Communications, Tokyo, Japan. He received his Ph.D. in Systems Science from Tokyo Institute of Technology in 1990. He was a Visiting Scientist in Computer Science at University of North Carolina at Chapel Hill in 1992 and 1993.
He received the Best Young Researchers Award from the Japan Society for Fuzzy Theory and Systems in 1990, the Outstanding Papers Award at the 1990 Annual NAFIPS Meeting in Toronto, Canada, in 1990, the Outstanding Papers Award at the Joint Hungarian-Japanese Symposium on Fuzzy Systems and Applications in Budapest, Hungary, in 1991, the Best Young Researchers Award from the Japan Society for Mechanical Engineers in 1994, the Outstanding Book Awards from the Japan Society for Fuzzy Theory and Systems in 1995, 1999 IFAC World Congress Best Poster Paper Prize in 1999, 2000 IEEE Transactions on Fuzzy Systems Outstanding Paper Award in 2000, the Best Paper Selection at 2005 American Control Conference in Portland, USA, in 2005, the SICE Award at RoboCup Japan Open 2010, Osaka, Japan, in 2010, Best in class Autonomy Award at RoboCup 2011 Japan Open in Osaka, Japan, in 2011, the Best Paper Award at 2013 IEEE International Conference on Control System, Computing and Engineering (ICCSCE 2013) in Penang, Malaysia, in 2013, the Best Paper Finalist at 2013 International Conference on Fuzzy Theory and Its Applications (iFUZZY2013) in Taipei, Taiwan, in 2013.
He served an Associate Editor for Automatica and for the IEEE Transactions on Fuzzy Systems. He served also as Chair of Task Forces on Fuzzy Control Theory and Application, IEEE Computational Intelligence Society Fuzzy Systems Technical Committee. He is currently on the IEEE Control Systems Society Conference Editorial Board.
His research interests include fuzzy systems control, nonlinear systems control, unmanned aerial vehicle, robotics, brain-machine interface and their applications. He published many papers in these areas, as well as 17 books, including: Fuzzy Control Systems Design and Analysis: A Linear Matrix Inequality Approach (Wiley-Interscience, 2001). His publications currently report over 18,000 citations according to Google Scholar Citations, with an h-index of 46 and an i10 index of 99. He is an IEEE fellow and an IFSA fellow.

Tutorial on Type-2 Fuzzy Sets and Systems

Organized by Christian Wagner, Jon Garibaldi and Robert John

Abstract

General type-2 fuzzy sets and systems are paradigms which enable fine-grained capturing, modelling and reasoning with uncertain information. While recent years have seen increasing numbers of applications from control to intelligent agents and environmental management, the perceived complexity of general type-2 fuzzy sets and systems still makes their adoption a daunting and not time-effective proposition to the majority of researchers.

This tutorial is designed to give researchers a practical introduction to general type-2 fuzzy sets and systems. Over three hours, the modular tutorial will address three main aspects of using and working with general type-2 fuzzy sets and systems:

  1. Introduction to General Type-2 Fuzzy Sets and Systems
    The first component of the tutorial will provide attendees with a concise and practice-led overview of general type-2 fuzzy sets and systems, reviewing the motivation behind their definition, their structure in relation to type-1 and interval type-2 fuzzy sets and systems, as well a set of recent applications.
  2. Designing General Type-2 Fuzzy Sets and Systems
    In the second part of the tutorial, two distinct aspects will be discussed. First, attendees will be given a practical introduction to designing their own general type-2 fuzzy system. Using the online browser-based toolkit JuzzyOnline, participants will be guided in the design of a general type-2 fuzzy system, relating their own design to the design of type-1 fuzzy systems at each stage.
    Second, the design of general type-2 fuzzy sets will be discussed through a presentation of a key set of recently introduced processes to create general type-2 fuzzy sets from data.
  3. Coding General Type-2 Fuzzy Sets and Systems
    The final part of the tutorial will focus on the programmatic implementation and use of general type-2 fuzzy sets and systems. Currently available software tools and toolkit for general type-2 fuzzy sets and system applications will be briefly reviewed, before interested participants will be supported in the development of a simple general type-2 fuzzy system based on the freely available Juzzy and/or R based general type-2 APIs.
Timing: The overall time of the tutorial will be three hours, with an approximately even split over all three tutorial components listed above (i.e. one hour per component).
Pre-requisites: basic knowledge of type-1 fuzzy sets and systems is the only pre-requisite for attendees to be able to benefit from this tutorial.

Biography

Christian Wagner

Dr Christian Wagner is an Associate Professor in Computer Science at the University of Nottingham, UK. He received his PhD in Computer Science from the University of Essex in 2009 after which he was involved both in the management and scientific work of the EU FP7 project ATRACO, joining the University of Nottingham in 2011. His main research interests are centred on uncertainty handling, approximate reasoning (reasoning in the face of uncertainty, lack of knowledge and vagueness), decision support and data-driven decision making using computational intelligence techniques. Recent applications of his research have focused in particular on decision support in environmental and infrastructure planning & management contexts as well as cyber-security. He has published more than 60 peer-reviewed articles in international journals and conferences, two of which recently won best paper awards (Outstanding IEEE Transactions on Fuzzy Systems paper 2013 (for a paper in 2010) and a best paper award for a Fuzz-IEEE 2012 conference paper), and several book chapters. Dr Wagner is currently active PI and Co-I on a number of research projects, with overall funding as PI of £1 million and funding as Co-I of £2 million. He is a senior member of the IEEE, an Associate Editor of the IEEE Transactions on Fuzzy Systems journal (IF: 6.3) and is actively involved in the academic community through for example the organization of special sessions at premiere IEEE conference such as the World Congress on Computational Intelligence 2014 and the IEEE Conference on System, Man and Cybernetics 2015. He has developed and been involved in the creation of multiple open source software frameworks, making cutting edge research accessible both to peer researchers as well as to different (multidisciplinary - beyond computer science) research and practitioner communities, including R and Java based toolkits for type-2 fuzzy systems in use in more than ten countries.

Jon Garibaldi

Prof. Jon Garibaldi is Head of the Intelligent Modelling and Analysis (IMA) Research Group in the School of Computer Science at the University of Nottingham. His main research interest is in developing intelligent techniques to model human reasoning in uncertain environments, with a particular emphasis on the medical domain. Prof. Garibaldi has been the PI on EU and EPSRC projects worth over £3m, and CoI on a portfolio of grants worth over £25m. He is Director of the University of Nottingham Advanced Data Analysis Centre, established in 2012 to provide leading-edge data analysis services across the University and for industrial consultancy. His experience of leading large research projects includes his roles as Lead Scientist and Co-ordinator of BIOPTRAIN, a Marie-Curie Early Stage Training network in bioinformatics optimisation worth over €2m, the local co-ordinator of the €6.4m BIOPATTERN FP6 Network of Excellence, lead Computer Scientist on a £700k MRC DPFS (Developmental Pathway Funding Scheme) project to transfer the Nottingham Prognostic Index for breast cancer prognosis into clinical use. Industrial projects include a TSB funded project for data analysis in the transport sector, and a collaborative project with CESG (GCHQ) investigating and modelling variation in human reasoning in subjective risk assessments in the context of cyber-security. He is currently the local PI for Nottingham on the £900k UKCRC Joint Funders Tissue Directory and Coordination Centre, a CoI on the £14m BBSRC/EPSRC Synthetic Biology Research Centre in Sustainable Routes to Platform Chemicals, and was CoI on the £10m BBSRC/EPSRC Centre for Plant Integrative Biology. Prof. Garibaldi has published over 200 articles on fuzzy systems and intelligent data analysis, including over 50 journal papers and over 150 conference articles, three book chapters, and three co-edited books. He is an Associate Editor of Soft Computing, was Publications Chair of FUZZ-IEEE 2007 and General Chair of the 2009 UK Workshop on Computational Intelligence, and has served regularly in the organising committees and programme committees of a range of leading international conferences and workshops, such as FUZZ-IEEE, WCCI, EURO and PPSN. He is a member of the IEEE.

Robert John

Prof. Robert John is a Professor of Operational Research and Computer Science and Head of the ASAP research group. He is a senior member of IEEE, fellow of the British Computer Society and elected member of the EPSRC college. In the field of type-2 fuzzy logic, his work is widely recognised by the international fuzzy logic community as leading in the aspects of theoretical foundations, as well as practical applications. His work has produced many fundamental new results that have opened the field to new research, enabling a broadening of scope and application. He is associate editor of the journal Soft Computing and the International Journal of Information & System Sciences, and member of the editorial board of International Journal of Cognitive Neurodynamics, Grey Systems: Theory and Application, Turkish Journal of Fuzzy Systems, International Journal for Computational Intelligence and Information and System Sciences. He chaired EUSFLAT2001 organised on behalf of the European Society of Fuzzy Logic and Technology and held at De Montfort University. He has over 150 publications of which circa 50 are in international journals and many papers are very well cited.

Computer Vision: A Computational Intelligence Perspective

Organized by Derek T. Anderson, Chee Seng Chan and James M. Keller

Abstract

We will discuss challenges in modern computer vision (CV) research and possible directions, tools and novel ideas that the computational intelligence (CI) community may contribute. We will discuss difficult problems or challenges in CV that is recognized by researchers in the areas of low-level, mid-level and high-level CV. We review standard and modern CV approaches, discuss data sets currently used by the CV community, and we present some CI techniques employed in each area, always with an eye towards where soft computing can make the best impact. This event is not meant to be a survey of all techniques; if yours is left out, please do not get mad. The intent is to provide an assessment, from our perspective, of the power, limitations, and potential of CI algorithms, with thoughts about the challenges for those of us in the CI family to have our technologies be better accepted by the CV community.

Below is a tentative list of topics and their subsequent organization

Part 1

Introduction:

  • Why study CV?
  • CV is steeped in probabilistic methods
  • Active research communities, e.g., PAMI, ICCV, CVPR, ECCV, NIPS, etc.
  • Comparison heavy (methods and data sets)
  • CV tools: OpenCV, SimpleCV and VLFeat
  • Provide open source code: educational and basis for benchmarking
  • David Marr: principles of least commitment and graceful degradation
Overview of Low/Mid/High Level CV:
  • Discussion of the boundaries and the fuzziness of them
  • Examples of the different levels and highlight some FS-based methods in each area
  • Ask the question of where does FSs fit in (offer the most pay off and/or make the most sense)?
Feature Learning with CNNs (Neural Networks) and iECO (Evolutionary Computation):
  • Why learn features and different approaches
  • Explain iECO
  • Go into signal/image processing part
  • Go into evolutionary search part
  • Explain CNNs
  • Go into CNN specifics
  • Go into applications for computer vision and explosive target detection
Fusion:
  • Explain fusion and why it is needed
  • Explain various levels
  • Explain fuzzy measures and fuzzy integrals
  • Extensions
  • Learning
  • Feature space fusion and kernels
  • Learning
  • Signal/image processing examples
Linguistic Summarization:
  • Demand (technically and application domains)
  • FS-based approaches
  • Linguistic summarization of video techniques
  • Metrics/measures that work on protoforms
Discussion:
  • Where does (should) CI belong in CV?
  • Low level: hard to press advantage of FS, not a difficult sale for NNs, EC everywhere
  • Medium: if case can be made for modeling the uncertainty for FSs, but likely NNs, again EC
  • High: biggest payoff for FSs because closest to human-like operations, NNs?, EC still
  • Need to demonstrate results on common (large) data sets
  • Need to identify new ways of evaluating the added benefit of FSs (not necessarily NNs) in CV
Part 2

Single Person Behavior/Activity Analysis:
  • Basic concepts
  • Features
  • Learning (zero-shot, one-shot and conventional learning)
  • Single image human motion analysis (HMA)
  • Video sequence HMA
  • Early event detection
  • Datasets
Multiple Person Behavior/Activity Analysis:
  • Basic concepts
  • Features
  • Learning
  • Datasets
Other Behavior/Activity Analysis (Crowd, Ego-motion):
  • Basic concepts
  • Features
  • Learning
  • Datasets
Discussion:
  • Applications and Challenges

Biography

Derek T. Anderson

Derek T. Anderson is an Assistant Professor in Electrical and Computer Engineering (ECE) at Mississippi State University (MSU). He received his B.S. and M.S. degrees in Computer Science and the Ph.D. in ECE. His research interests are new frontiers in data and information fusion for pattern analysis and automated decision making with an emphasis on heterogeneous uncertain information. This includes measure theory and fuzzy integrals, clustering, multi-source (sensor, algorithm and human) fusion, remote sensing and computer vision. Derek received the Best Student Paper Award at FUZZ-IEEE 2008, the Best Paper Award at FUZZ-IEEE 2012 and was a co-author of the Best Student Paper Award at SPIE in ATR in 2013. He has received funding from DARPA, Airforce Research Laboratory (AFRL), National Institute of Justice (NIJ), Leonard Wood Institute (LWI), Pacific Northwest National Laboratory (under a U.S. Department of Energy contract), Army Research Office (ARO) and NVESD (Countermine and Science and Technology) and U.S. Army ERDC. He has published 1 book chapter, 18 journal manuscript, 57 conference proceedings and he is an Associate Editor for the IEEE Transactions on Fuzzy Systems (TFS). Derek has co-chaired numerous special sessions at WCCI and IPMU. He also ran a workshop (in conjunction with Jim Keller and Tony Han) on the “View of Computer Vision Research and Challenges for the Fuzzy Set Community”, FUZZ-IEEE 2013.

Chee Seng Chan

Chee Seng Chan is a Senior Lecturer in the Faculty of Computer Science and Information Technology, University of Malaya, Malaysia. He received his PhD from University of Portsmouth, U.K. in 2008. His research interests spans a variety of aspects of fuzzy qualitative reasoning and computer vision; with a focus on image/video content analysis and human-robot interaction. He is the founder chair for the IEEE Computational Intelligence Society (CIS) Malaysia chapter and founder of the Malaysian Image Analysis and Machine Intelligence Association (MIAMI), a society under the International Association of Pattern Recognition (IAPR). He is/was the organizing chair for the Asian Conference on Pattern Recognition (ACPR) in 2015, general chair for the IEEE Visual Communications and Image Processing (VCIP) in 2013, and has co-chaired numerous special sessions at FUZZ-IEEE (2010-2015). Also, he has served as the guest editor in International Journal of Uncertainty, Fuzziness and Knowledge-based Systems (IJUFKS), Information Sciences (INS) and Signal, Video and Image Processing (SVIP). He is a recipient of the Hitachi Research Fellowship in 2013 and the IET (Malaysia) Young Engineer award in 2010. Finally, he is a senior member of IEEE, a chartered engineer and member of IET.

James M. Keller

James M. Keller holds the University of Missouri Curators Professorship in the Electrical and Computer Engineering and Computer Science Departments on the Columbia campus. He is also the R. L. Tatum Professor in the College of Engineering. His research interests center on computational intelligence with a focus on problems in computer vision, pattern recognition, and information fusion including bioinformatics, spatial reasoning, geospatial intelligence, landmine detection and technology for eldercare. Professor Keller has coauthored over 400 technical publications. Jim is a Life Fellow of the IEEE, an IFSA Fellow, and past President of NAFIPS. He received the 2007 Fuzzy Systems Pioneer Award and the 2010 Meritorious Service Award from the IEEE Computational Intelligence Society. He finished a full six year term as Editor-in-Chief of the IEEE Transactions on Fuzzy Systems, followed by being the Vice President for Publications of the IEEE CIS from 2005-2008, and since then an elected CIS Adcom member. He is the IEEE TAB Transactions Chair and a member of the IEEE Publication Review and Advisory Committee. Jim has had many conference positions and duties over the years.


Dynamic Fuzzy Neural Networks: Architectures, Algorithms and Applications

Organized by Meng Joo Er

Abstract

It is well known fuzzy logic provides human reasoning capabilities to capture uncertainties, which cannot be described by precise mathematical models. In essence, a fuzzy logic system is a rule-based system, which comprises a set of linguistic rules in the form of “ IF-THEN”. Designing a fuzzy system is a subjective approach, which is adopted to express a designer's knowledge. As there is no formal and effective way of knowledge acquisition, it is difficult for a designer, even he/she is a domain expert, to examine all the input-output data from a complex system so as to find a number of appropriate rules for the fuzzy system. In order to circumvent this problem, it is desirable to develop an objective approach to automate the modeling process based on numerical training data for fuzzy systems.

Neural networks offer remarkable advantages, such as adaptive learning, parallelism, fault tolerance, and generalization. They have been proved to be powerful techniques in the discipline of system control, especially when the controlled system is difficult to be modeled accurately, or the controlled system has large uncertainties and strong nonlinearities. Thus, fuzzy logic and neural networks have been widely adopted in model-free adaptive control of nonlinear systems resulting in neural-network-based fuzzy systems termed Fuzzy Neural Network (FNN) Systems.

Usually, the typical approaches of designing FNNs are to build standard neural networks, which are designed to approximate a fuzzy algorithm or a process of fuzzy inference through the structure of neural networks. These FNNs can readily solve two problems of conventional fuzzy reasoning: 1) The lack of systematic design for membership functions because it is basically a heuristic approach, and 2) The lack of adaptability for possible changes in the reasoning environment. The two problems intrinsically concern parameter estimation. Yet, in most FNNs, structure identification is still time-consuming because the determination of hidden nodes in neural networks can be viewed as the choice of number of fuzzy rules. On the other hand, we can see that most of the existing FNNs are trained by the Back-Propagation (BP) algorithm. It is well known that the BP method is generally slow and likely to become trapped in local minima. Hence, a fast learning paradigm for real-time applications is highly desirable.

Dynamic Fuzzy Neural Networks (DFNN) are FNN systems whose structure is evolving and/or self-organising. Some researchers call DFNN as Self-organisng FNN or evolving FNN. The objectives of this tutorial are to review Dynamic Fuzzy Neural Networks developed by various researchers over the last few decades. It will have a comprehensive coverage of three aspects of DFNN, namely Architectures, Algorithms, and Applications. There are numerous kinds of FNN proposed in the literatur and most of them are suitable for only off-line learning. Online learning algorithms are more attractive as they can be used for online identification and control processes in dealing with real-world engineering problems which are nolinear, time-varying and ill-defined dynamic systems. In this tutorial, both offline and online methods will be presented.

Structure of the Tutorial

  1. Review of Fuzzu Logic and Neural Networks
  2. Artchitecture of the DFNN
  3. Learning Algorithms of DFNN
    • Offline learning vs online learning
    • A review of state of the arts me
    • Dynamic Fuzzy Neural Network (D-FNN) based on extended Radial Basis Function (RBF) neural networks, which are functionally equivalent to Takagi-Sugeno-Kang (TSK) fuzzy systems.
    • Dynamic Fuzzy Neural Network (ED-FNN) learning algorithm, which inherits some of the attractive features of the original D-FNN algorithm.
    • Generalized Dynamic Fuzzy Neural Network (GD-FNN) based on Ellipsoidal Basis Function (EBF), which implement TSK fuzzy inference systems, are presented to extract fuzzy rules from input-output sample patterns.
  4. Real-world Applications of DFNN
    • Postsurgical Blood Pressure Regulation
    • Classification of Breast Cancer
    • Mobile Robots
    • Unmanned Ground Vehicles
    • Unmanned Aerial Vehicles
    • Adaptive Noise Cancellation

Biography

Professor Er Meng Joo is currently a Full Professor in Electrical and Electronic Engineering, Nanyang Technological University, Singapore. He served as the Founding Director of Renaissance Engineering Programme and an elected member of the NTU Advisory Board and from 2009 to 2012. He served as a member of the NTU Senate Steering Committee from 2010 to 2012.
He has authored five books entitled “Dynamic Fuzzy Neural Networks: Architectures, Algorithms and Applications” and “Engineering Mathematics with Real-World Applications” published by McGraw Hill in 2003 and 2005 respectively, and “Theory and Novel Applications of Machine Learning” published by In-Tech in 2009, “New Trends in Technology: Control, Management, Computational Intelligence and Network Systems” and “New Trends in Technology: Devices, Computer, Communication and Industrial Systems”, both published by SCIYO, 19 book chapters and more than 500 refereed journal and conference papers in his research areas of interest.
Professor Er was bestowed the Web of Science Top 1 % Best Cited Paper and the Elsevier Top 20 Best Cited Paper Award in 2007 and 2008 respectively. In recognition of the significant and impactful contributions to Singapore’s development by his research projects, Professor Er won the Institution of Engineers, Singapore (IES) Prestigious Engineering Achievement Award twice (2011 and 2015) and the highest honour of the IES Innovation Challenge 2015. Under his leadership, the NTU Team emerged first runner-up in the Freescale Technology Forum Design Challenge 2008. He is also the only dual winner in Singapore IES Prestigious Publication Award in Application (1996) and IES Prestigious Publication Award in Theory (2001). He received the Teacher of the Year Award for the School of EEE in 1999, School of EEE Year 2 Teaching Excellence Award in 2008, the Most Zealous Professor of the Year Award in 2009 and the Outstanding Mentor Award in 2014. He also received the Best Session Presentation Award at the World Congress on Computational Intelligence in 2006 and the Best Presentation Award at the International Symposium on Extreme Learning Machine 2012. On top of this, he has more than 50 awards at international and local competitions.
Currently, Professor Er serves as the Editor-in-Chief of Transactions on Machine Learning and Artificial Intelligence and the International Journal of Electrical and Electronic Engineering and Telecommunications. He also serves an Area Editor of International Journal of Intelligent Systems Science and an Associate Editor of 11 refereed international journals, namely IEEE Transaction on Fuzzy Systems, IEEE Transaction on Cybernetics, International Journal of Fuzzy Systems, Neurocomputing, ETRI Journal, Journal of Robotics, International Journal of Applied Computational Intelligence and Soft Computing, International Journal of Fuzzy and Uncertain Systems, International Journal of Automation and Smart Technology, International Journal of Modelling, Simulation and Scientific Computing, International Journal of Intelligent Information Processing and an editorial board member of the EE Times.
Professor Er has been invited to deliver more than 60 keynote speeches and invited talks overseas. He has also been active in professional bodies. He has served as Chairman of IEEE Computational Intelligence Society (CIS) Singapore Chapter (2009 to 2011) and Chairman of IES Electrical and Electronic Engineering Technical Committee (EEETC) (2004 to 2006 and 2008 to 2012). Under his leadership, the IEEE CIS Singapore Chapter won the CIS Outstanding Chapter Award 2012 (The Singapore Chapter is the first chapter in Asia to win the award). In recognition of his outstanding contributions to professional bodies, he was bestowed the IEEE Outstanding Volunteer Award (Singapore Section) and the IES Silver Medal in 2011. Due to his outstanding contributions in education, research, administration and professional services, he is listed in Who’s Who in Engineering, Singapore, Edition 2013.

FIND US ON

line line

IEEE CEC 2016 Tutorials

Meta-heuristics for Multi-objective Optimization

Organized by Carlos Coello

This tutorial provides with a general picture of the current state-of-the-art in multiobjective optimization using metaheuristics. First, some historical background is provided, dating back to the origins of multiobjective optimization in general. This discussion motivates the use of metaheuristics for solving multiobjective problems and includes a brief description of some of the earliest approaches proposed in the literature. Then, a discussion on different heuristics used for multiobjective optimization is provided. This discussion includes evolutionary algorithms, simulated annealing, tabu search, scatter search, the ant system, particle swarm optimization and artificial immune systems. The tutorial finishes with a discussion of some of the research topics that seem more promising for the next few years.

Biography

Carlos Artemio Coello Coello received a PhD in Computer Science from Tulane University (USA) in 1996. He is currently full professor with distinction at CINVESTAV-IPN in Mexico City, Mexico. He has published over 400 papers in international peer-reviewed journals, book chapters, and conferences. He has also co-authored the book "Evolutionary Algorithms for Solving Multi-Objective Problems", which is now in its Second Edition (Springer, 2007) and has co-edited the book "Applications of Multi-Objective Evolutionary Algorithms" (World Scientific, 2004). His publications currently report over 28,000 citations, according to Google Scholar (his h-index is 67). He received the "2007 National Research Award" (granted by the Mexican Academy of Science) in the area of "exact sciences" and, since January 2011, he is an "IEEE Fellow" for "contributions to multi-objective optimization and constraint-handling techniques."

He is also the recipient of the prestigious "2013 IEEE Kiyo Tomiyasu Award" and of the "2012 National Medal of Science and Arts" in the area of "Physical, Mathematical and Natural Sciences" (this is the highest award that a scientist can receive in Mexico). He also serves as associate editor of the IEEE Transactions on Evolutionary Computation, Computational Optimization and Applications, Pattern Analysis and Applications, Journal of Heuristics, Evolutionary Computation and Applied Soft Computing. He has served as Vice-Chair and Chair of the IEEE CIS Evolutionary Computation Technical Committee and is currently the Chair of the IEEE CIS Distinguished Lecturers Committee. He was also the General Chair of the 2013 IEEE Congress on Evolutionary Computation, which took place in Cancún, Mexico.

Applying Evolutionary Computation in Industrial Practice

Organized by Erik D. Goodman

Abstract

Many researchers have developed evolutionary algorithms that excel in solving various types of problems that would seem to make them attractive for companies to use on such problems. In experience, though, it is difficult to get the staff of a company even to entertain such an idea. This tutorial will explore some factors that make it difficult and some keys to success, in either a consulting/sponsored research relationship to industry or in founding a company to provide search/optimization products or services. Time will be provided for addressing questions from the participants. The presenter co-founded Red Cedar Technology, Inc. (one such successful company), directed a university-based design center and an industrial consortium, and did many industry-sponsored R&D projects at the university. He now directs BEACON, an NSF-sponsored Science and Technology Center that includes about 20 faculty members studying computational evolution in various forms including GA/GP/ES and digital evolution.

Biography

Erik D. Goodman is PI and Director of the BEACON Center for the Study of Evolution in Action, an NSF Science and Technology Center headquartered at Michigan State University, funded at $47.5 million for 2010-2020. BEACON now has over 600 members, including many who study evolutionary computation or digital evolution. Goodman studies application of evolutionary principles to solution of engineering design problems. He received the Ph.D. in computer and communication sciences from the University of Michigan in 1972. He joined MSU’s faculty in Electrical Engineering and Systems Science in 1971, was promoted to full professor in 1984, and also holds appointments in Mechanical Engineering and in Computer Science and Engineering, in which he has guided many Ph.D. students. He directed the Case Center for Computer-Aided Engineering and Manufacturing from 1983-2002, and MSU’s Manufacturing Research Consortium from 1993-2003. He co-founded MSU’s Genetic Algorithms Research and Applications Group (GARAGe) in 1993. In 1999, he co-founded Red Cedar Technology, Inc., which develops design optimization software, and was Vice President for Technology until BEACON was founded in 2010. The company was sold in 2013 and continues to operate. Goodman was chosen Michigan Distinguished Professor of the Year, 2009, by the Presidents Council, State Universities of Michigan. He was Chair of the Executive Board and a Senior Fellow of the International Society for Genetic and Evolutionary Computation, 2003-2005. He was founding chair of the ACM’s SIG on Genetic and Evolutionary Computation (SIGEVO) in 2005.

Automatic Algorithm Configuration: Methods, Applications, and Perspectives

Organized by Thomas Stützle

Abstract

The design of algorithms for computationally hard problems is time-consuming and difficult. This is in large part due to a number of aggravating circumstances such as the NP-hardness of most of the problems to be solved, the difficulty of algorithm analysis due to stochasticity and heuristic biases, and the large number of degrees of freedom in defining and selecting algorithmic components and settings of numerical parameters. Even when using off-the-shelf solvers, their performance strongly depends on the appropriate settings of a large number of parameters that can influence their search behaviour. Over the recent years, automatic algorithm configuration methods have been developed to effectively search large and diverse parameter spaces for identifying superior algorithm designs and performance improving parameter settings. These methods have by now proved to be instrumental for developing high-performance algorithms.

In the first part of this tutorial, I will introduce the algorithm design and tuning tasks that recent automatic algorithm configuration methods address; describe the main existing automatic algorithm configuration techniques; and show how (easily) they can be used. In the second part of the tutorial, I will discuss various successful applications of automatic algorithm configuration methods to configure mixed-integer programming solvers, to generate hybrid stochastic local search algorithms, to design multi-objective optimisers, and to improve algorithm anytime behaviour. Finally, we will discuss specific aspects relevant to the application of automatic algorithm configuration methods and argue that automatic algorithm configuration methods will transform the way algorithms for difficult problems are designed and developed in the future.

Biography

Dr. Stützle is a senior research associate of the Belgian F.R.S.-FNRS working at the IRIDIA laboratory of Université libre de Bruxelles (ULB), Belgium. He received the Diplom (German equivalent of M.S. degree) in business engineering from the Universität Karlsruhe (TH), Karlsruhe, Germany in 1994, and his PhD and his habilitation in computer science both from the Computer Science Department of Technische Universität Darmstadt, Germany in 1998 and 2004, respectively. He is the co-author of two books about ``Stochastic Local Search: Foundations and Applications'' and ``Ant Colony Optimization'' and he has extensively published in the wider area of metaheuristics including 19 edited proceedings or books, 9 journal special issues, and more than 170 peer-reviewed articles, many of which are highly cited. He is associate editor of Applied Mathematics and Computation, Computational Intelligence, and Swarm Intelligence and on the editorial board of five other journals including Evolutionary Computation and JAIR. His main research interests are in metaheuristics, swarm intelligence, methodologies for engineering stochastic local search algorithms, multi-objective optimization, and automatic algorithm configuration. In fact, since more than a decade he is interested in automatic algorithm configuration and design methodologies and he has contributed to several effective algorithm configuration techniques such as F-race, Iterated F-race and ParamILS. His 2002 GECCO paper on "A Racing Algorithm For Configuring Metaheuristics" (joint work with M. Birattari, L. Paquete, and K. Varrentrapp) has received the 2012 SIGEVO impact award.

Advances in Particle Swarm Optimization

Organized by Andries Engelbrecht

Abstract

The main objective of this tutorial will be to answer the question if particle swarm optimization (PSO) can be considered as a universal optimizer. In the context of this tutorial, this means that the PSO can be applied to a wide range of optimization problem types as well as search domain types. The tutorial will start with a very compact overview of the original, basic PSO. Some experience and background on PSO will be assumed. A summary of important theoretical findings about PSO, in particular particle trajectories and convergence behavior will be provided, as this will provide important insights to the remainder of the tutorial. This will be followed by a short discussion on heuristics to select proper values for control parameters. The remainder and bulk of the tutorial will cover a classification of different problem types, and will show how PSO can be applied to solve problems of these types. This part of the tutorial will be organized in the following sections, one for each problem type:

  • Continuous-valued versus discrete-valued domains
  • Unimodal versus multi-modal landscapes
  • Multi-solution problems requiring niching capabilities
  • Constrained versus unconstrained problems, also covering boundary constraints
  • Multi-objective optimization
  • Dynamic environments
  • Dynamic Multi-objective optimization
  • Optimization with dynamically changing constraints
For each problem type, it will be shown why the standard PSO can not solve these types of problems efficiently. Simple adaptations to the PSO that will allow it to solve each problem type will then be discussed. The focus will be on PSO adaptations that do not violate the foundational principles of PSO. For each of these problem types a small subset of the most successful algorithms will be discussed.

Biography

Andries Engelbrecht received the Masters and PhD degrees in Computer Science from the University of Stellenbosch, South Africa, in 1994 and 1999 respectively. He is Professor in Computer Science at the University of Pretoria, and serves as Head of the department. He holds the position of South African Research Chair in Artificial Intelligence, and leads the Computational Intelligence Research Group. His research interests include swarm intelligence, evolutionary computation, neural networks, artificial immune systems, and the application of these paradigms to data mining, games, bioinformatics, finance, and difficult optimization problems. He has published over 270 papers in these fields and is author of two books, Computational Intelligence: An Introduction and Fundamentals of Computational Swarm Intelligence.

Prof Engelbrecht is very active in the international community, annually serving as reviewer for over 20 journals and 10 conferences. He is an Associate Editor of the IEEE Transactions on Evolutionary Computation, Journal of Swarm Intelligence, IEEE Transactions on Computational Intelligence and AI in Games, and Soft Computing. He was co-guest editor of special issues of the IEEE Transactions on Evolutionary Computation and the Journal of Swarm Intelligence. He served on the international program committee and organizing committee of a number of conferences, organized special sessions, presented tutorials, and took part in panel discussions. He was the founding chair of the South African chapter of the IEEE Computational Intelligence Society. He is a member of the Evolutionary Computation Technical Committee, Games Technical Committee, and the Evolutionary Computation in Dynamic and Uncertain Environments Task Force.

Search Based Software Engineering: Foundations, Recent Advances, Challenges and Future Research Directions

Organized by Marouane Kessentini

Abstract

Software engineering is by nature a search problem to find an optimal or near-optimal solution. This search is often complex with several competing constraints, and conflicting functional and non-functional objectives. The situation can be worse since nowadays successful software are more complex, more critical and more dynamic leading to an increasing need to automate or semi-automate the search process of acceptable solutions for software engineers. As a result, an emerging research area, called Search-Based Software Engineering (SBSE), is rapidly growing. SBSE is a software development practice which focuses on couching software engineering problems as optimization problems and utilizing meta-heuristic and computational search techniques to discover and automate the search of near optimal solutions to those problems.

SBSE has been applied to wide variety of software engineering problems covering the software life cycle activities such as testing, requirements engineering, software management, refactoring, re-modularization, etc. While SBSE has been successfully applied to a wide variety of software engineering problems, several challenges are still to be addressed. I will focus in this talk on covering the basic concepts related to the formulation of large scale real world software engineering problems as search problems such as test cases generation, model transformation, code refactoring, etc. Then, I will describe several successful SBSE projects in automotive industry and give several future research directions to handle the growing scalability issues when adapting computational search and intelligence techniques to real world software engineering problems.

Biography

Dr. Marouane Kessentini is an Assistant Professor in the Department of Computer and Information Science at the University of Michigan. He is the founder of the Search-Based Software Engineering (SBSE) research lab including now one post-doc, six PhD students and seven master students. He has several collaborations with different industrial companies on studying software engineering problems by optimization techniques such as software quality, software testing, software migration, software evolution, etc. He also received the best dissertation award in 2012 from University of Montreal and a Presidential BSc Award from the President of Tunisia in 2007. He published more than 70 papers in software engineering conferences and journals including 3 best paper awards. He has served as program committee member in several major conferences (GECCO, MODELS, ICMT, SSBSE, etc.) and as organization member of many conferences and workshops. He wa also the co-chair of the SBSE track at the GECCO2014/2015 conferences and the general chair of of the Search Based Software Engineering Symposium (SSBSE2016). He is the founder of the North American Symposium on Search Based Software Engineering.

Differential Evolution with Ensembles and Topologies

Organized by P. N. Suganthan and M. Z. Ali

Abstract

Differential Evolution (DE) is one of the most powerful stochastic real-parameter optimization algorithms of current interest. DE operates through similar computational steps as employed by a standard Evolutionary Algorithm (EA). However, unlike traditional EAs, the DE-variants perturb the current-generation population members with the scaled differences of distinct population members. Therefore, no separate probability distribution has to be used for generating the offspring. Since its inception in 1995, DE has drawn the attention of many researchers all over the world resulting in a lot of variants of the basic algorithm with improved performance. This tutorial will begin with a brief overview of the basic concepts related to DE, its algorithmic components and control parameters. It will subsequently discuss some of the significant algorithmic variants of DE for bound constrained single-objective optimization. Recent modifications of the DE family of algorithms for multi-objective, constrained, large-scale, niching and dynamic optimization problems will also be included. The talk will discuss the effects of incorporating ensemble learning in DE – a novel concept that can be applied to swarm & evolutionary algorithms to solve various kinds of optimization problems. The talk will also discuss neighborhood topologies based DE to improve the performance of DE on multi-modal landscapes. Theoretical advances made to understand the search mechanism of DE and the effect of its most important control parameters will be discussed. The talk will finally highlight a few problems that pose challenge to the state-of-the-art DE algorithms and demand strong research effort from the DE-community in the future.

Biography

Ponnuthurai Nagaratnam Suganthan received the B.A degree, Postgraduate Certificate and M.A degree in Electrical and Information Engineering from the University of Cambridge, UK in 1990, 1992 and 1994, respectively. After completing his PhD research in 1995, he served as a pre-doctoral Research Assistant in the Dept of Electrical Engineering, University of Sydney in 1995–96 and a lecturer in the Dept of Computer Science and Electrical Engineering, University of Queensland in 1996–99. He moved to NTU in 1999. He is an Editorial Board Member of the Evolutionary Computation Journal, MIT Press. He is an associate editor of the IEEE Trans on Cybernetics (2012 - ), IEEE Trans on Evolutionary Computation (2005 - ), Information Sciences (Elsevier) (2009 - ), Pattern Recognition (Elsevier) (2001 - ) and Int. J. of Swarm Intelligence Research (2009 - ) Journals. He is a founding co-editor-in-chief of Swarm and Evolutionary Computation (2010 - ), an Elsevier Journal. SaDE paper (published in April 2009) won "IEEE Trans. on Evolutionary Computation" outstanding paper award in 2012. Dr Jane Jing Liang (his former PhD student) won the IEEE CIS Outstanding PhD dissertation award, in 2014. IEEE CIS Singapore Chapter won the best chapter award in Singapore in 2014 for its achievements in 2013 under his leadership. His research interests include swarm and evolutionary algorithms, pattern recognition, numerical optimization by population-based algorithms and applications of swarm, evolutionary & machine learning algorithms. His publications have been well cited (Googlescholar Citations: 17k). His SCI indexed publications attracted over 1000 SCI citations in each calendar years 2013, 2014 and 2015. He was selected as one of the highly cited researchers by Thomson Reuters in 2015 in computer science. He served as the General Chair of the IEEE SSCI 2013. He has been a member of the IEEE (S'90-M'92-SM'00-F'15) since 1990 and an elected AdCom member of the IEEE Computational Intelligence Society (CIS) in 2014-2016

Niching Methods for Multimodal Optimization

Organized by Michael Epitropakis and Xiaodong Li

Abstract

Population or single solution search-based optimization algorithms (i.e. {meta,hyper}-heuristics) in their original forms are usually designed for locating a single global solution. Representative examples include among others evolutionary and swarm intelligence algorithms. These search algorithms typically converge to a single solution because of the global selection scheme used. Nevertheless, many real-world problems are "multimodal" by nature, i.e., multiple satisfactory solutions exist. It may be desirable to locate many such satisfactory solutions, or even all of them, so that a decision maker can choose one that is most proper in his/her problem domain. Numerous techniques have been developed in the past for locating multiple optima (global and/or local). These techniques are commonly referred to as “niching” methods. A niching method can be incorporated into a standard search-based optimization algorithm, in a sequential or parallel way, with an aim to locate multiple globally optimal or suboptimal solutions,. Sequential approaches locate optimal solutions progressively over time, while parallel approaches promote and maintain formation of multiple stable sub-populations within a single population. Many niching methods have been developed in the past, including crowding, fitness sharing, derating, restricted tournament selection, clearing, speciation, etc. In more recent times, niching methods have also been developed for meta-heuristic algorithms such as Particle Swarm Optimization, Differential Evolution and Evolution Strategies.

In this tutorial we will aim to provide an introduction to niching methods, including its historical background, the motivation of employing niching in EAs. We will present in details a few classic niching methods, such as the fitness sharing and crowding methods. We will also provide a review on several new niching methods that have been developed in meta-heuristics such as Particle Swarm Optimization and Differential Evolution.  Employing niching methods in real-world situations still face significant challenges, and this tutorial will discuss several such difficulties. In particular, niching in static and dynamic environments will be specifically addressed. Following this, we will present a suite of new niching function benchmark functions specifically designed to reflect the characteristics of these challenges. Performance metrics for comparing niching methods will be also presented and their merits and shortcomings will be discussed. Experimental results across both classic and more recently developed niching methods will be analysed based on selected performance metrics. Apart of benchmark niching test  functions, several examples of applying niching methods to solving real-world optimization problems will be provided. This tutorial will use several demos to show the workings of niching methods.

Biography

Xiaodong Li received his B.Sc. degree from Xidian University, Xi'an, China, and Ph.D. degree in information science from University of Otago, Dunedin, New Zealand, respectively. Currently, he is an Associate Professor at the School of Computer Science and Information Technology, RMIT University, Melbourne, Australia. His research interests include evolutionary computation, neural networks, complex systems, multiobjective optimization, and swarm intelligence. He serves as an Associate Editor of the IEEE Transactions on Evolutionary Computation, the journal of Swarm Intelligence (Springer), and International Journal of Swarm Intelligence Research. He is a founding member and currently a Vice-chair of the following three IEEE CIS Task Forces: Swarm Intelligence, Large Scale Global Optimization, and Multimodal Optimization. He was the General Chair of SEAL'08, a Program Co-Chair AI'09, and a Program Co-Chair for IEEE CEC’2012. He is the recipient of 2013 SIGEVO Impact Award. For further information, please visit his website.

Theory of Evolutionary Algorithms

Organized by Benjamin Doerr and Carola Doerr

Abstract

This tutorial provides a smooth introduction to the theory of evolutionary algorithms. We shall start by discussing

  • what theory of evolutionary algorithms is and aims at,
  • how to read and interpret results from the theory literature,
  • some central results and insight of the last 20 years.
In a second part of the tutorial, we provide an introduction to the most successful subarea of theory, namely runtime analysis, which aims at understanding how long an evolutionary algorithm takes to find a solution of a certain quality. We also discuss the complementing concept of black-box complexity, a notion designed to capture the limits of commonly applied evolutionary algorithms. Finally, we give an example of how to use the insights from the theory of evolutionary computation to design novel and more efficient evolutionary algorithms.

Biography

Benjamin Doerr

Benjamin Doerr is a full professor at the French Ecole Polytechnique. He also is an adjunct professor at Saarland University. He received his diploma (1998), PhD (2000) and habilitation (2005) in mathematics from Kiel University. His research area is the theory both of problem-specific algorithms and of randomized search heuristics like evolutionary algorithms. Major contributions to the latter include runtime analyses for evolutionary algorithms and ant colony optimizers, as well as the further development of the drift analysis method, in particular, multiplicative and adaptive drift. In the young area of black-box complexity, he proved several of the current best bounds.

Together with Frank Neumann and Ingo Wegener, Benjamin Doerr founded the theory track at GECCO, served as its co-chair 2007-2009 and 2014. He is the Hot-off-the-press chair for GECCO 2016. He is a member of the editorial boards of "Evolutionary Computation", "Natural Computing", "Theoretical Computer Science" and "Information Processing Letters". Together with Anne Auger, he edited the book "Theory of Randomized Search Heuristics". He gave tutorials on various theory topics at GECCO, CEC, PPSN, and other venues.

Carola Doerr

Carola Doerr is a permanent CNRS researcher at the Université Pierre et Marie Curie (Paris 6). She studied mathematics at Kiel University (Germany, Diploma in 2007) and computer science at the Max Planck Institute for Informatics and Saarland University (Germany, PhD in 2011). From Dec. 2007 to Nov. 2009, Carola Doerr has worked as a business consultant for McKinsey & Company, mainly in the area of network optimization, where she has used randomized search heuristics to compute more efficient network layouts and schedules. Before joining the CNRS she was a post-doc at the Université Diderot (Paris 7) and the Max Planck Institute for Informatics.

Carola's main research interest is in the theory of randomized algorithms, both in the analysis of existing algorithms as well as in the design of novel algorithmic approaches. She has published several papers about black-box complexity and has contributed to the field of evolutionary computation also through results on the runtime analysis of evolutionary algorithms and drift analysis, as well as through the development of search heuristics for solving geometric discrepancy problems. Carola has been chairing the theory track at GECCO in 2015 (together with Francisco Chicano) and she serves as tutorial chair of PPSN 2016 (together with Nicolas Bredeche). Since 2014 she is also involved in the organization of the women@GECCO workshop, where she is chairing the organization committee in 2016. She has been a tutorial speaker at GECCO 2013, 2014, and 2016.

FIND US ON

line line

General Information

The IEEE WCCI 2016 will feature workshops to be held in conjunction with the main event. The overall purpose of a workshop is to provide participants with the opportunity to present and discuss novel research ideas on active and emerging topics of Computational Intelligence.

IEEE WCCI 2016 Workshops

  • Workshop-01 The 3rd International Workshop on Computational Energy Management in Smart Grids
  • Workshop-02 International Workshop on Neuromorphic Computing and Cyborg Intelligence
  • Workshop-03 Key Challenges and Future Directions of Evolutionary Computation
  • Workshop-04 Computational Intelligence Approaches for Multi-view Data Analytics

IEEE WCCI 2016 Workshops


Workshop-01 The 3rd International Workshop on Computational Energy Management in Smart Grids

Organized by Stefano Squartini, Derong Liu, Francesco Piazza, Dongbin Zhao and Haibo He

Workshop-01A - Monday, 25 July 2016, 8:00 - 10:00am, Room: 208-209
Workshop-01B - Tuesday, 26 July, 4:30 - 6:30pm, Room: 208-209

The sustainable usage of energy resources is actually an issue that humanity and technology have been seriously facing in the last decade, as a consequence of the higher and higher energy demand worldwide and the strong dependence on oil-based fuels. This shoved the scientists and technicians worldwide to intensify their studies on renewable energy resources, especially in the Electrical Energy sector. At the same time, a remarkable increment of the complexity of the electrical grid has been also registered at diverse levels in order to include variegated and distributed generation and storage sites, resulting in strong engineering challenges in terms of energy distribution, management and system maintenance. This yielded in a flourishing scientific literature on sophisticated algorithms and systems aimed at introducing intelligence within the electrical energy grid with several effective solutions already available in the market. These efforts have also recently cross-fertilized both research and development of commercial products for other grid types, as the smart water and natural gas grids, which have been registering an increasing interest in the last five years.

The many different needs coming from heterogeneous grid customers, at diverse grid level, and the different peculiarities of energy sources to be included in the grid itself, makes the task challenging and multi-faceted. Along this same direction, a big variety of interventions can be applied into the grid to increase the inherent degree of automation, optimal functioning, security and reliability, thus increasing the engineering appeal of the issue. A multi-disciplinary coordinated action is therefore required to the scientific communities operating in the Electrical and Electronic Engineering, Computational Intelligence, Digital Signal Processing and Telecommunications research fields to provide adequate technological solutions, having in mind the more and more stringent constraints in terms of environmental sustainability.

Focalizing to the interests of our scientific community, the organizers of this Workshop wants to explore the new frontiers and challenges within the Computational Intelligence research area, including Neural Networks based solutions, for the optimal usage and management of energy resources in Smart Grid scenarios. Indeed, the recent adoption of distributed sensor networks in many grid contexts enabled the availability of data to be used to develop suitable expert systems with the aim of supporting the humans in dealing with the complex problems in grid management, from multiple applicative perspectives. Related research is undoubtedly already florid, but many open issues need to be studied and innovative intelligent systems investigated.

By moving from the success obtained by the CEMiSG2014 Workshop organized within the IJCNN2014 conference in Beijing (China) and by the CEMiSG2015 Workshop organized within the IJCNN2015 conference in Killarney (Ireland), the third edition of the CEMiSG Workshop is still targeted to propose a proficient discussion table for scientists joining the IJCNN2016 conference at the WCCI2016.

Topics

Workshop topics include, but are not limited to:

  • Computational Intelligence for Smart Grids Applications
  • Neural Networks based algorithms for Complex Energy Systems
  • Soft Computing based Algorithms in Energy Applications
  • Expert Systems for Smart Grid Optimization
  • Smart Grids and Big Data
  • Computational Intelligence for Vehicle to Grid
  • Automatic Fault Detection Algorithms in Smart Grid scenarios
  • Computational methods for Smart Grid Self-Healing
  • Learning-based Control of Renewable Energy Generators
  • Smart Building Energy Management
  • Deep Neural Networks for Energy Efficiency
  • Computational Intelligence for Energy Internet Management
  • Energy Resource Allocation and Task Scheduling
  • Short/Long-term Load Forecasting
  • Demand-side Management
  • Learning Systems for Smart AMIs
  • Neural Networks for Time Series Prediction in Smart Grids
  • Non-Intrusive Load Monitoring
  • Hybrid Battery Management
Click here for more information and website of the workshop.


Workshop-02 International Workshop on Neuromorphic Computing and Cyborg Intelligence

Organized by Huajin Tang, Gang Pan, Arindam Basu and Luping Shi

Workshop-02 - Tuesday, 26 July, 8:00 - 10:00am, Room: 208-209

Emulating brain-like learning performance has been a key challenge for research in neural networks and learning systems, including recognition, memory and perception. In the last few decades, a wealth of machine learning approaches have been proposed including sparse representations, hierarchical and deep learning neural networks. While achieving impressive performance these methods still compare poorly to biological systems and the problem of reducing the amount of human supervision and computations needed for learning remains a challenge.

On the other hand, the development of novel data representation and learning approaches from recent advances in neuromorphic systems have shown appealing computational advantages. For example, using neural coding theory to represent the external sensory data, and developing spiking timing based learning algorithm have achieved real-time learning performance, either in neuromorphic computational models or hardware systems.  Attributed to the new visual or auditory sensors, neuromorphic hardware has provided a fundamentally different technique for data representation, i.e., asynchronous events rather than frames of images as in main stream recognition algorithms. However, the current neuromorphic information processing algorithms are not comparable to achieve sophisticated features and power learning performance as what machine learning approaches can offer. One promising method is to develop integrated learning models that apply brain-like data presentation and learning mechanisms, e.g., implementing deep learning in neuromorphic systems. Neuromorphic systems also overlap with another framework called cyborg intelligence, combining brain functions with computational machines to achieve the best of both via brain-machine interface. The workshop will target the challenging problems in these areas by reporting new solutions, theoretical and technical advances in neuromorphic computing and cyborg intelligence from the worldwide researchers and engineers.

Relevant Topics

Workshop topics include, but are not limited to:

  • Cognitive computing and cyborg intelligence
  • Neuromorphic information/signal processing
  • Brain-inspired data representation models
  • Neuromorphic learning and cognitive systems
  • Spike-based sensing and learning
  • Neuromorphic sensors and hardware systems
  • Intelligence for embedded systems
  • Cognition mechanisms for big data
  • Embodied cognition and neuro-robotics


Workshop-03 Key Challenges and Future Directions of Evolutionary Computation

Organized by Yun Li, Cesare Alippi, Thomas Bäck, Piero Bonissone, Stefano Cagnoni, Carlos Coello Coello, Oscar Cordón, Kalyanmoy Deb, David Fogel, Marouane Kessentini, Yuhui Shi, Xin Yao and Mengjie Zhang

Workshop-03A - Wednesday, 27 July 2016, 2:30 - 4:30pm, Room: 208-209
Workshop-03B - Wednesday, 27 July 2016, 4:30 - 6:30pm, Room: 208-209

Since the first WCCI taking place in Orlando in 1994, this Congress series and the Evolutionary Computation community have progressed tremendously. A number of CEC Panel Sessions were held and explored future directions of Evolutionary Computation. As part of the forthcoming WCCI, CEC 2016 in Vancouver promises c.50 Special Sessions, covering comprehensive activities. A Workshop to explore "Key Challenges and Future Directions of Evolutionary Computation" and to reach consensus among academia and industry is therefore timely. This format (instead of a discussion-only forum or panel session) will allow position papers that are submitted, peer reviewed and duly accepted to be recorded in the CEC Workshop Proceedings for future references.

Following individual presentations of accepted position papers, small breakout sessions will be held in parallel to explore deeper and broader views. Then an open panel discussion will proceed for convergence among academia and industry. It is intended that a short summary will be written up later for IEEE Computational Intelligence Magazine as a separate article for the CIS community and beyond.

Relevant Topics

You are warmly invited to submit a position paper with rationale, rigour and supporting evidence on one or more of the following aspects

  • key challenges in Evolutionary Computation (such as NP-complexity, constraint handling, convergence proof, convergence vs robustness, optimality vs robustness, multi-/many-objectives, benchmarks, benchmarking problems);
  • future directions of Evolutionary Computation (such as multi-modal optimisation, non-stationary evolution, multi-rate or learning evolution, variable-size or fuzzy evolution, distributed and concurrent optimisation, cloud-based evolution, automated design and customisation, optimal evolution, predictive evolution, hybrid ecological and cultural evolution, quantum and analogue 'evolutionary computation');
  • real-world challenges to applications of Evolutionary Computation (such as application demand vs supply, real-world NP-hardest problems, landscape forecasting, complexity reduction, generational evolution vs interactive learning, real-time issues, streaming evolution, ease of implementation, scalability and parallelism, big data application);
  • and their emergence and trending behaviours (such as Thomson Reuters' "Hot Topics", "Research Fronts" and "Essential Science Indicators").

Technical Requirements

You position paper should be academic and should normally contain

  • An introduction - clearly identifying the issue and your position, and written in a way that catches the reader's attention
  • A perspective - based on critical reviews, and providing facts for a solid foundation of your arguments
  • A statement of your position or the challenge - formulating and limiting your chosen issue or argument carefully; supporting or validating your position through inductive reasoning with statistical data, authoritative references, or interviews with industrialist experts; analysing opportunities and threats
  • A discussion of both sides of the issue - examining the strengths and weaknesses of your position and potential alternatives
  • Perceived milestones - suggesting courses of action that you have deduced for future developments or in addressing their challenges
  • Potential impacts - evaluating possible solutions and their impacts on the subject, the field and the wider world
  • A conclusion - summarising and reinforcing (without repeating the introduction or body of the paper) the main points, millstones and impacts that you have formulated and advocated.


Workshop-04 Computational Intelligence Approaches for Multi-view Data Analytics

Organized by Giovanni Montana, Carlo Francesco Morabito and Roberto Tagliaferri

Workshop-02 - Tuesday, 26 July, 8:00 - 10:00am, Room: 208-209

Over the years, huge quantities of data have been generated by large-scale scientific experiments (biomedical, “omic”, imaging, astronomical, etc.), big industrial companies and on the web. One of the characteristics of such Big Data is that it often includes multiple “views” of the underlying objects. For instance, in biomedical research, various “omics” technologies (e.g. mRNA, miRNA etc.) or imaging technologies (MRI, CT, PET, etc.) can generate multiple measurements of the same biological samples.

In order to exploit this level of complexity, new machine learning and computational intelligence methodologies such as neural networks and graph mining, amongst others, have been proposed to analyze and/or visualize these multi-facets datasets in an attempt to better exploit the information they contain.

The aim of the workshop is to solicit new approaches to real world scientific and industrial big data integration.

Relevant Topics

Papers must present original work or review the state-of-the-art in the following non-exhaustive list of topics:

  • Multi-view learning
  • data fusion
  • data integration
  • multi-view and multi domain data applications
  • Multiview data visualization
Click here for more information.

FIND US ON

line line

Panel Sessions



Panel sessions bring together the perspectives of many panelists into a cohesive conversation of innovative ideas. These panel sessions will consist of presentations/open discussions by panelists who will directly engage with the conference audience. These sessions add an enriching dimension to the conference experience and a welcome networking alternative to traditional paper presentations, which dominate some conferences.

IEEE WCCI 2016 Panel Sessions


IEEE WCCI 2016 is pleased to confirm a Panel Session on “Turning Big Data Challenges into Opportunities”

Date: 25 July 2016 (Monday)
Time: 2:30 - 4:30 pm
Venue: Ballroom B

Background image

Xin Yao (Moderator)
2014-2015 IEEE CIS President

Background image

Kathy Grise (Moderator)
IEEE Future Directions Program Director

Background image

Andrew Feng
Yahoo

Background image

Dimitar Filev
Ford

Background image

Danil Prokhorov
Toyota

Background image

Yuandong Tian
Facebook

Background image

Piero Bonissone
Piero P Bonissone Analytics LLC


IEEE WCCI 2016 is pleased to confirm a Panel Session on “How to Publish Your Research Papers in IEEE CIS Transactions?”

Date: 26 July 2016 (Tuesday)
Time: 2:30 - 4:30 pm
Venue: Ballroom B

Background image

Haibo He
Editor-in-Chief, IEEE Transactions on Neural Networks and Learning Systems

Background image

Chin-Teng Lin
Editor-in-Chief, IEEE Transactions on Fuzzy Systems

Background image

Kay Chen Tan
Editor-in-Chief, IEEE Transactions on Evolutionary Computation

Background image

Graham Kendall
Editor-in-Chief, IEEE Transactions on Computational Intelligence and AI in Games

Background image

Yaochu Jin
Editor-in-Chief, IEEE Transactions on Cognitive and Developmental Systems

Background image

Hisao Ishibuchi
Editor-in-Chief, IEEE Computational Intelligence Magazine

Background image

Jacek M. Zurada
IEEE Life Fellow, Past IEEE VP-Technical Activities

Background image

Nikhil R. Pal (Moderator)
IEEE CIS Vice-President for Publications


IEEE WCCI 2016 is pleased to confirm a Panel Session on “IEEE and CIS in the Next Decade”

Date: 27 July 2016 (Wednesday)
Time: 2:30 - 4:30 pm
Venue: Ballroom B

Background image

Karen Bartleson
2016 IEEE President-elect

Background image

Vincenzo Piuri
2015 IEEE VP - Technical Activities

Background image

Jacek M. Zurada
2014 IEEE VP - Technical Activities

Background image

Xin Yao
2014-2015 IEEE CIS President

Background image

Marios M. Polycarpou
2012-2013 IEEE CIS President

Background image

Pablo A. Estévez
Moderator, 2016-2017 IEEE CIS President


IEEE WCCI 2016 is pleased to confirm a Panel Session on “The IEEE CIS History Panel: Yesterday, Today, and Tomorrow”

Date: 26 July 2016 (Tuesday)
Time: 4:30 - 6:30 pm
Venue: Ballroom B

Background image

Piero Bonissone (Moderator)
President of NNS, 2002

Background image

James Bezdek
President of NNS, 1997 - 98

Background image

Russell Eberhart
President of NNS, 1992 - 93

Background image

Vincenzo Piuri
President of CIS, 2006 - 07

Background image

Rudolf Seising
Historian of Fuzzy Logic and CIS

Background image

Jacek M. Zurada
President of CIS, 2004 - 05


A memorial session will be organized at IEEE WCCI 2016 to commemorate the lifetime achievements of David Casasent (1942-2015).

This session addresses issues of advanced image processing, optical image processing, and pattern recognition, with an emphasis of recent advances in Big Data. The session includes invited talk and a panel covering these topics in the light of the work of David Casasent. The featured speaker for the Dave Casasent memorial session will be Prof. Ashit Talukder from the University of North Carolina at Charlotte.

Date: 27 July 2016 (Wednesday)
Time: 4:30 - 5:25 pm
Venue: Ballroom B

Background image

Ali Minai (Moderator)

Background image

Hava Siegelmann (Moderator)

Background image

Robert Kozma (Moderator)


A memorial session will be organized at IEEE WCCI 2016 to commemorate the lifetime achievements of Walter J. Freeman (1927-2016).

Computational neurodynamics is a field which has been significantly defined and shaped by the pioneering work of Walter J. Freeman (1927-2016). His pivotal contributions include the discovery and modeling of the nonlinear dynamical oscillations utilized by vertebrate brains to create perception. Recent experimental and theoretical advances in these areas will be discussed in this session in a talk and by panelists.

Date: 27 July 2016 (Wednesday)
Time: 5:35 - 6:30 pm
Venue: Ballroom B

Background image

Ali Minai (Moderator)

Background image

Hava Siegelmann (Moderator)

Background image

Robert Kozma (Moderator)

FIND US ON

line line

General Information

The IEEE WCCI 2016 will feature competitions to be held in conjunction with the main event. The overall purpose of a workshop is to provide participants with the opportunity to present and discuss novel research ideas on active and emerging topics of Computational Intelligence.

Prospective competition organizers are invited to submit their proposals to the Competitions Chair, Dr. Chang-Shing Lee and Dr. Simon M. Lucas by 15 December 2015. A common website for all accepted competitions will be provided on the congress website at www.wcci2016.org, for which a link to the homepage of each competition maintained by their organizers independently will be provided.

IEEE WCCI 2016 Competitions

  • Competition-01 Competition on Bound Constrained Single Objective Numerical Optimization
  • Competition-02 Human vs. Computer Go Competition
  • Competition-03 AutoML Challenge II: Fully Automatic Machine Learning without ANY human intervention
  • Competition-04 International Time Series Forecasting Competition Computational Intelligence in Forecasting (CIF)
  • Competition-05 General Video Game AI (GVG-AI) Competition
  • Competition-06 The Numenta Anomaly Benchmark Competition For Real Time Anomaly Detection
  • Competition-07 The Niching Methods for Multimodal Optimization
  • Competition-08 FML-based Applications to Social Media Competition
  • Competition-09 Multiobjective Optimization Competition
  • Competition-10 IEEE Grand Challenge on Flickr YFCC100M User Tag and Caption Prediction

Competitions Program

  1. Sunday, 24 July, 2016
    • 9:00 - 11:00am
      Competition-02: Human vs. Computer Go Competition (Layout and Test the Internet)
    • 12:30 - 5:00pm
      Competition-02: Human vs. Computer Go Competition

  2. Monday, 25 July, 2016
    • 8:00am - 6:30pm
      Competition-02: Human vs. Computer Go Competition

  3. Tuesday, 26 July, 2016
    • 8:00 - 11:30am
      Competition-02: Human vs. Computer Go Competition (Go Forum)
      Competition-08: FML-based Applications to Social Media Competition
    • 2:30 - 4:30pm
      Competition-01: Bound Constrained Single Objective Numerical Optimization
      Competition-03: AutoML Challenge II: Fully Automatic Machine Learning without ANY human intervention
    • 4:30 - 6:30pm
      Competition-04: International Time Series Forecasting Competition Computational Intelligence in Forecasting (CIF)
      Competition-05: General Video Game AI (GVG-AI) Competition

  4. Wednesday, 27 July, 2016
    • 8:00 - 10:00am
      Competition-06: The Numenta Anomaly Benchmark Competition For Real Time Anomaly Detection
      Competition-07: The Niching Methods for Multimodal Optimization
    • 2:30 - 4:30pm
      Competition-09: Multiobjective Optimization Competition
      Competition-10: IEEE Grand Challenge on Flickr YFCC100M User Tag and Caption Prediction
    • 4:30 - 6:30pm
      Competitions 01-10: Discussion and Report

IEEE WCCI 2016 Competitions


Competition on Bound Constrained Single Objective Numerical Optimization

Organized by P N Suganthan, Mostafa Z. Ali, Qin Chen, J. J. Liang and B. Y. Qu

Competition goals

The goals are to evaluate the current state of the art in single objective optimization with bound constraints and to propose novel benchmark problems with diverse characteristics. The algorithms will be evaluated with very small number of function evaluations to large number of function evaluations as well as single solution to multiple solutions. Under the above scenarios, novel problems will be designed for the first time to emulate real-world problem solving. In particular the following cases would also be considered:

  1. An industry is interested in solving different instances of the same class of problem. How can we learn from past instances to solve the future instances more effectively?
  2. 2. For a given single objective problem, locate N top solutions separated from each other by a specified Euclidean distance.
These two cases have not been considered much in the evolutionary computation community, although these scenarios are commonly encountered in real-world problem solving settings.

Contributions to the Evolutionary Computation Community

Single objective numerical optimization is the most important class of problems. All new evolutionary and swarm algorithms are tested on single objective benchmark problems. In addition, these single objective benchmark problems can be transformed into dynamic, niching composition, computationally expensive and many other classes of problems.

How to submit an entry and how to evaluate them

Potential authors are asked to make use of the software of benchmark problems (in Matlab, C or Java) to be distributed from the competitions web pages to test their algorithms either with or without surrogate methods. The authors have to execute their novel or existing algorithms on the given benchmark problems and present the results in various formats as outlined in the technical report. The evaluation criteria will also be specified in the technical report. The authors are asked to prepare a conference paper detailing the algorithms used and the results obtained on the given benchmark problems and submit their papers to the associated special session within CEC 2016. The authors presenting the best results should also be willing to release their software for verification before declaring the eventual winners of the competition.

Special Session Associated With This Competition

This competition requires all entries to have an associated conference paper submitted. We also expect at least one author of each entry to register, attend the conference and present their paper(s).


Human vs. Computer Go Competition

Organized by Chang-Shing Lee, Shi-Jim Yen, I-Chen Wu and Hsin-Hung Chou

Competition goals

In order to enhance the fun in playing Go between humans and computer Go programs and to stimulate the development and researches of computer Go programs, we submit this proposal and hope to have a chance to hold Human vs. Computer Go Competition @ IEEE WCCI 2016, Vancouver Convention Centre, Vancouver, Canada. The level of computer Go program in 19x19 approaches one human with 6D before Google AlphaGo came out in Oct. 2015. In addition, the rank of Google AlphaGo with more than 1000 CPUs and 100 GPUs is 9P in Mar. 2016. Hence, we will focus on human vs. computer Go competition under small computational hardware resource. The objective of the competition is to highlight an ongoing research on Computational Intelligence approaches as well as their future applications on game domains.

Expected Humans

  • Chun-Hsun Chou (9P), Taiwan
  • Ping-Chiang Chou (6P), Taiwan
  • Shang-Rong Tsai (6D), Taiwan
  • Shi-Jim Yen (6D), Taiwan
  • Sheng-Shu Chang (6D), Taiwan

Expected Computer Go Programs

  • Zen (Japan)
  • Dark Forest/ Facebook (USA)
  • CGI (Taiwan)


AutoML Challenge II: Fully Automatic Machine Learning without ANY human intervention

Organized by Hugo Jair Escalante, ChaLearn and Codalab

Abstract

Enter the final rounds of the AutoML challenge, with prizes donated by Microsoft and NVIDIA: create a fully automatic learning machine capable of solving classification and regression tasks without any human intervention. We launched AutoML in 2015 as part of the IJCNN 2015 competition program. Rounds 0 through 2 have been completed and the challenge presently entered round 3 (advanced). The challenge will terminate in March 2016. New this year: we added a GPU track sponsored by NVIDIA, which should reinforce the possibilities for deep learning methods (a.k.a. neural networks) to contribute. Winning is not as hard as you may think: nobody beat the baseline method using Naive Bayes in the last AutoML phase! Simple methods are sometimes best.


International Time Series Forecasting Competition Computational Intelligence in Forecasting (CIF)

Organized by Martin Štěpnička and Michal Burda

Competition Method

The prediction competition is open to all methods of computational intelligence, incl. fuzzy method, artificial neural networks, evolutionary algorithms, decision & regression tress, support vector machines, hybrid approaches etc. used in all areas of forecasting, prediction & time series analysis, etc. Ensemble techniques are also allowed, if they employ any CI method. The contestants will use a unique consistent methodology for all the time series. The only evaluation criterion is the Symmetric Mean Absolute Percentage Error (SMAPE). The results will be uncovered during the IEEE WCCI 2016.


General Video Game AI (GVG-AI) Competition

Organized by Diego Perez and Simon Lucas

Abstract

For WCCI 2016, we propose to run a two-player planning version (track – we call each main variant of GVG-AI a track). For this track, each player is given access to a forward model of the game, but do not know in advance what game is being played, nor of course what the opponent will do. This is a logical next step in the challenge, and we expect it to create a good deal of interest (and we will keep the sponsorship of Google DeepMind).


The Numenta Anomaly Benchmark Competition For Real Time Anomaly Detection

Organized by Alexander Lavin and Subutai Ahmad

Competition Goal

Much of the world’s data is streaming, time-series data, where anomalies give significant information in often-critical situations. Yet detecting anomalies in streaming data is a difficult task, requiring detectors to process data in real-time, not batches, and learn while simultaneously making predictions. The Numenta Anomaly Benchmark (NAB) attempts to provide a controlled and repeatable environment of open-source tools to test and measure anomaly detection algorithms on streaming data.

The NAB competition will distribute awards for two distinct components:

  1. Algorithms: What is the best algorithm for detecting anomalies in streaming data? Participants interested in this component will submit anomaly detection algorithms and the results of running their algorithms on NAB.
  2. Dataset: What is the best data contribution to the benchmark corpus? Participants interested in this component will submit real-world, time-series data with labeled anomalies for inclusion into NAB.
Submissions are due July 1, 2016 to allow the judges ample time to validate the results. Yet we encourage early submissions and questions such that we may help ensure the submissions qualify. Please send submissions and questions to us here.

For more information, including rules and submission details, visit this contest webpage.


The Niching Methods for Multimodal Optimization

Organized by Michael G. Epitropakis, Xiaodong Li and Andries Engelbrecht

Competition Aim

The aim of the competition is to provide a common platform that encourages fair and easy comparisons across different niching algorithms. The competition allows participants to run their own niching algorithms on 20 challenging benchmark multimodal functions with different characteristics and levels of difficulty. Researchers are welcome to evaluate their niching algorithms using this benchmark suite, and report the results by submitting a paper to the associated niching special session (i.e., submitting via the online submission system of CEC'2016).


FML-based Applications to Social Media Competition

Organized by Chang-Shing Lee, Giovanni Acampora,, Ryosuke Saga, Marek Reformat and Hsin-Hung Chou

Competition Topic

“Who will like your article that you posted on Facebook?” Please design a Fuzzy Markup Language (FML) system to predict how many likes in your posted article within one to three weeks. Competitors have to describe which variables are involved in the knowledge base (KB) of FML system. Competitors can use an expert-based or a machine learning approach to identify the rule base.

Competition Method

Competition will be done before the conference. We will release the Java-based FML tool and call for applications to construct the knowledge base and rule base of FML. They should construct the FML system and write system description document with 2 or 3 pages. The competition will be held on the Internet. The winners can be invited to present the FML system at the IEEE WCCI 2016.

Details about the competition can be found here.


Multiobjective Optimization Competition

Organized by Hui Li, Qingfu Zhang, P. N. Suganthan, Aimin Zhou, Kalyanmoy Deb, Hisao Ishibuchi and Carlos Coello Coello

Over the past thirty years, multiobjective evolutionary algorithms (MOEAs) have become the prominent methodologies for solving multiobjective optimization problems (MOPs). The major strength of MOEAs is that they are able to find an approximation of the whole Pareto front (PF) in a single run. Fitness assignment and diversity maintenance are two major research issues in MOEAs. To design an efficient MOEA, the balance between them must be carefully considered. It should be mentioned that none of the MOEAs can have good performance on all MOPs. Therefore, it is quite important to investigate the suitability of MOEAs on the problem difficulties. During the past a few years, the following problem features that can challenge MOEAs in convergence or diversity have attracted much attention:

  • the geometric PF and PS shapes
  • the biased search space
  • the degeneracy of PF
  • the multimodality
  • the high dimensionalities of variable space and objective space.
This competition is devoted to compare the performance of existing or new MOEAs on a set of 22 unconstrained challenging benchmark test problems with the above-mentioned problem features.


IEEE Grand Challenge on Flickr YFCC100M User Tag and Caption Prediction

Organized by Bart Thomee, Pierre Garrigues, Liangliang Cao and David A. Shamma

Challenge overview

The members of the Flickr community manually tag photos with the goal of making them searchable and discoverable. With the advent of mobile phone cameras and auto-uploaders, photo uploads have become more numerous and asynchronous, and manual tagging is cumbersome for most users. Progress has been largely driven by training deep neural networks on datasets, such as ImageNet, that were built by manual annotators. However, acquiring annotations is expensive. In addition, the different categories of annotations are defined by researchers and not by users, which means they are not necessarily relevant to users’ interests, and cannot be directly leveraged to enable search and discovery.

We believe our proposed challenge fills a void that is not currently addressed by existing challenges. Our challenge is uniquely aligned with the context of computational intelligence and machine learning: (1) our dataset contains on the order of 100 million photos, which reflects well the challenges of understanding multimedia at large scale, and (2) our benchmark focuses on user-generated content, where a large vocabulary of concepts is collected from tags annotated by users. In contrast, the ImageCLEF annotation task focuses on the occurrence of 251 English words that appear on web pages instead of being directly associated with images, while the ImageNet challenge considers synsets from the WordNet dictionary as annotations, and many of them do not appear in real-world images like those uploaded to Flickr.

Our challenge focuses on replicating how people annotate photos, rather than just focusing on photo annotation without the human component. Our challenge asks participants to build image analysis systems that think like humans:the correct annotation for an image isn’t necessarily the “true label”. For example, while a photo containing an apple, a banana and a pear could be annotated using these three words, a person more likely would annotate the image with the single word “fruit”.

As the problem of automatic image annotation is not close to being solved, we intend to hold our grand challenge during multiple years. Depending on the progress of the submissions and the state of the art, the difficulty of the challenge could increase.

FIND US ON

line line

General Information

Apart from attending the technical programs, participants are also invited to attend various social events to be held during IEEE WCCI 2016, such as Welcome Reception, Award Banquet, Women in Computational Intelligence (WCI) Reception, Student Activities and Young Professionals Reception, IEEE CIS Chapters Forum, etc.

Welcome Reception

IEEE WCCI 2016 invites everyone to attend the congress Welcome Reception to be held at Ballroom D, Foyer, Terrace, Vancouver Convention Centre (West Building) on Sunday evening (7:00 - 9:00 pm), 24 July 2016.

FIND US ON

line line

Award Banquet

The congress Award Banquet will be held at WEST Exhibit Hall A, Vancouver Convention Centre (West Building) on Wednesday evening (7:30 - 10:00 pm), 27 July 2016.

FIND US ON

line line

CIS Activites

Women in Computational Intelligence (WCI) Reception

The Women in Computational Intelligence (WCI) Committee invites all the ladies to attend the WCI-Reception at Room 114-115, Vancouver Convention Centre (West Building) on Thursday evening (7:00 - 8:30 pm), 28 July 2016.

It is one of our traditions to organize receptions for women at the IEEE Computational Intelligence Society (CIS) conferences.

IEEE CIS is fully committed to ensure equal opportunities to both genders in the society's life and the computational intelligence arena and the WCI-committee develops, promotes and runs activities directed to achieve this goal.

Help us in building a strong WCI-Community!

Sanaz Mostaghim
Chair of WCI-Committee 2016

Student Activities and Young Professionals Reception

The IEEE Computational Intelligence Student Activities and Young Professionals sub-committees invites all students, young professionals, and their supervisors and mentors to attend a reception at Room 301-305, Vancouver Convention Centre (West Building) on Monday evening (7:30 - 9:00 pm), 25 July 2016.

The reception will be opened by Professor Pablo A. Estevez, President of the IEEE Computational Intelligence Society (CIS). This will be followed by a short presentation about opportunities and activities for students and young professionals within the CIS, and then a social networking activity (speed-dating).

Come along and network, meeting old friends and new ones!

Dipti Srinivasan
IEEE WCCI 2016 Student Activities Chair

Keeley Crockett
Chair of IEEE Computational Intelligence Student Activities

Albert Lam
Chair of IEEE Computational Intelligence Young Professionals

IEEE CIS Chapters Forum

The IEEE CIS has more than 100 Chapters and Student Chapters, composed of active volunteers who add value to the membership of our members by organizing events, lectures, summer schools, workshops, competitions as well as networking and social events.

We would like to invite all IEEE CIS Chapter Chairs/Officer volunteers, including Student Chapters officers, to the 2016 IEEE CIS Chapters Forum which will be organized as part of IEEE WCCI 2016 at Room 306, Vancouver Convention Centre (West Building) on Monday evening (7:30 - 9:00 pm), 25 July 2016. Members who are interested in setting up new IEEE CIS Chapters are also welcome to join the Forum.

During the Forum, we will discuss how to make the most out of the initiatives organized by the IEEE, how to receive funding to organize various events and to be informed on best practices.

As the number of places is limited, please send us an email expressing your interest in joining the Forum: demetrios.g.eliades@ieee.org

Demetrios Eliades
Chair of IEEE CIS Chapters Subcommittee

FIND US ON

line line

Technical Programs

IEEE WCCI 2016 Program - At - A - glance


Click here to view the IEEE WCCI 2016 Program - At - A - Glance.


IEEE WCCI 2016 Final Program Booklet


Click here to download the IEEE WCCI 2016 Final Program Booklet.

FIND US ON

line line