October 30, 2017

CEC Sessions

CEC-1  Special Session on Many-Objective Optimization
CEC-2  Special Session on Memetic Computing
CEC-3  Special Session on Evolutionary Robotics
CEC-4  Special Session on New Directions in Evolutionary Machine Learning
CEC-5  Special Session on Evolutionary Computation for Feature Selection, Extraction and Dimensionality Reduction
CEC-6  Special Session on Evolutionary Computing Application in Hardware
CEC-7  Special Session on Data-Driven Evolutionary Optimization of Computationally Expensive Problems
CEC-8  Special Session on Transfer Learning in Evolutionary Computation
CEC-9  Special Session on Evolutionary Computer Vision
CEC-10  Special Session on Evolutionary Computation for Service-Oriented Computing
CEC-11  Special Session on Advanced Evolutionary Computation Approaches for Smart Grid and Sustainable Energy Systems

CEC-1  Special Session on Many-Objective Optimization

Organized by Ran Cheng (ranchengcn@gmail.com), Miqing Li, Rui Wang, Xin Yao

The field of multi-objective optimization has developed rapidly over the last 20 years, but the design of effective algorithms for addressing problems with more than three objectives (called many-objective optimization problems, MaOPs) remains a great challenge. First, the ineffectiveness of the Pareto dominance relation, which is the most important criterion in multi-objective optimization, results in the underperformance of traditional Pareto-based algorithms. Also, the aggravation of the conflict between convergence and diversity, along with increasing time or space requirement as well as parameter sensitivity, has become key barriers to the design of effective many-objective optimization algorithms. Furthermore, the infeasibility of solutions’ direct observation can lead to serious difficulties in algorithms’ performance investigation and comparison. All of these suggest the pressing need of new methodologies designed for dealing with MaOPs, new performance metrics and test functions tailored for experimental and comparative studies of many-objective optimization algorithms.

Scope and Topics

We welcome high-quality original submissions addressing various topics related to many-objective optimization, but are not limited to:

  • Algorithms for many-objective optimization, including search operators, mating selection, environmental selection and population initialization;
  • Performance indicators for many-objective optimization;
  • Benchmark functions for many-objective optimization;
  • Visualization techniques for many-objective optimization;
  • Objective reduction techniques for many-objective optimization;
  • Preference articulation and decision making methods for many-objective optimization;
  • Constraint handling methods for many-objective optimization;
  • Many-objective optimization in combinatorial/discrete problems;
  • Many-objective optimization in dynamic environments;
  • Many-objective optimization in large-scale problems.

CEC-2  Special Session on Memetic Computing

Organized by Liang Feng (liangf@cqu.edu.cn), Chuan-Kang Ting, Maoguo Gong, Meng-Hiot Lim.

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
  • Evolutionary multi-tasking

CEC-3  Special Session on Evolutionary Robotics

Organized by Renan C. Moioli (moioli@isd.org.br), Patricia A. Vargas, Micael Couceiro, Josh Bongard, 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 centers all over the world. ER techniques are mostly inspired by existing biological architectures and Darwin’s principle of selective reproduction of the fittest. Evolution has revealed that living creatures are able to accomplish complex tasks required for their survival, thus embodying cooperative, competitive and adaptive behaviors.

Having an intrinsic interdisciplinary character, ER has been employed towards 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 theoretical analysis to real-life applications.

Scope and Topics

  • Evolution of robots that display minimal cognitive behavior, learning, memory, spatial cognition, adaptation 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 behavior.
  • Evolution of morphology in close interaction with the environment, giving rise to self-reconfigurable, 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-4  Special Session on New Directions in Evolutionary Machine Learning

Organized by Yusuke Nojima (nojima@cs.osakafu-u.ac.jp), Will Browne, Masaya Nakata

Evolutionary Machine Learning (EML) explores technologies that integrate machine learning (e.g., neural network, decision tree, fuzzy systems, reinforcement learning) with evolutionary computation for tasks including optimization, classification, regression, and clustering. Since machine learning contributes to parameter learning while evolutionary computation contributes to model/parameter optimization, 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 Genetics-Based Machine Learning (GBML) and it was concerned with learning classifier systems (LCS) with its numerous implementations. More recently, EML has emerged as a more general field than GBML. It is consequently a broader, more flexible and more capable paradigm than GBML.

Scope and Topics

The aim of this special session is to explore potential EML technologies and clarify new directions for EML to show its prospects. 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 neural network (e.g., neuroevolution, evolutionary deep neural network)
  • Evolutionary decision tree
  • Evolutionary cascade systems
  • Evolutionary fuzzy systems
  • Evolutionary reinforcement learning
  • Evolutionary ensemble systems
  • Evolutionary adaptive systems
  • Artificial immune systems
  • Genetic programming applied to machine learning
  • Evolutionary feature selection and construction for machine learning
  • Transfer learning; learning blocks of knowledge (memes, code, etc.)
  • Accuracy-interpretability tradeoff in EML
  • Applications and theory of EML

CEC-5  Special Session on Evolutionary Computation for Feature Selection, Extraction and Dimensionality Reduction

Organized by Bing Xue (bing.xue@ecs.vuw.ac.nz), Yaochu Jin, and Mengjie Zhang

In machine learning and data mining, the quality of the input data determines the quality of the output (e.g. accuracy), known as the GIGO (Garbage In, Garbage Out) principle. For a given problem, the input data of a learning algorithm is almost always expressed by a number of features (attributes or variables). Therefore, the quality of the feature space is a key for the success of any machine learning and data algorithm.

Feature selection, feature extraction or construction and dimensionality reduction are important and necessary data pre-processing steps to increase the quality of the feature space, especially with the trend of big data. Feature selection aims to select a small subset of important (relevant) features from the original full feature set. Feature extraction or construction aims to extract or create a set of effective features from the raw data or create a small number of (more effective) high-level features from (a large number of) low-level features. Dimensionality reduction aims to reduce the dimensionality of the data space with the focus of solving “the curse of dimensionality” issue. All of them can potentially improve the performance of a learning algorithm significantly in terms of the accuracy, increase the learning speed, and the complexity and the interpretability of the learned models.  However, they 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 these tasks due to the fast development of evolutionary computation and capability of stochastic search, constraint handling and dealing with multiple conflict objectives.

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 selection, feature extraction and construction, dimensionality reduction and related studies on improving quality of the feature space and their applications. Authors are invited to submit their original and unpublished work to this special session.  Topics of interest include but are not limited to:

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

CEC-6  Special Session on Evolutionary Computing Application in Hardware

Organized by Andy M Tyrrell (andy.tyrrell@york.ac.uk), Martin A Trefzer

Evolvable systems encompass understanding, modeling 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 robotics and evolutionary many-core system.

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 bio-inspired computing and hardware. Headline topics are evolvable systems techniques, bio-inspired computation with materials and engineering physical devices, evolutionary many-core systems and evolutionary robotics. The special session invites submissions in any of the following areas:

  • Hardware system optimisation
  • Learning and adaptation
  • Cooperation and competition
  • Co-evolution of robot morphologies
  • Real-world applications
  • Self-reconfigurable systems
  • Adaptive systems
  • Self- repairing systems
  • Fault-tolerant systems
  • Autonomous systems
  • Specialised hardware
  • Computational hardware and materials
  • Self-adaptation
  • Self-monitoring
  • Self-testing

CEC-7  Special Session on Data-Driven Evolutionary Optimization of Computationally Expensive Problems

Organized by Chaoli Sun, Jonathan Fieldsend, Yew-Soon Ong

Meta-heuristic algorithms, including evolutionary algorithms and swarm optimization, face challenges when solving time-consuming problems, as typically these approaches require thousands of function evaluations to arrive at solutions that are of reasonable quality. Surrogate models, which are computationally cheap, have in recent years gained popularity in assisting meta-heuristic optimization, by replacing the compute-expense/time-expensive problem during phases of the heuristic search. However, due to the curse of dimensionality, it is very difficult, if not impossible to train accurate surrogate models. Thus, appropriate model management techniques, memetic strategies, and other schemes are often indispensable. In addition, modern data analytics involving advance sampling techniques and learning techniques such as semi-supervised learning, transfer learning, and active learning are highly beneficial for speeding up evolutionary search while bringing new insights into the problems of interest.

Scope and Topics

This special session aims at bringing together researchers from both academia and industry to explore future directions in this field. The topics of this special session include but are not limited to the following topics:

  • Surrogate-assisted evolutionary optimization for computationally expensive problems
  • Adaptive sampling using machine learning and statistical techniques
  • Surrogate model management in evolutionary optimization
  • Data-driven optimization using big data and data analytics
  • Knowledge acquisition from data and reuse for evolutionary optimization
  • Computationally efficient evolutionary algorithms for large scale and/or many-objective optimization problems
  • Real world applications including multidisciplinary optimization

CEC-8  Special Session on Transfer Learning in Evolutionary Computation

Organized by Bing Xue (Bing.Xue@ecs.vuw.ac.nz), Liang Feng, Yew Soon Ong, Mengjie Zhang

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 generalize 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, including Genetic algorithms (GAs), Genetic programming (GP), Evolutionary programming (EP), Evolution strategies (ES), Learning classifier systems (LCS), Particle swarm optimization (PSO), Ant colony optimization (ACO), Differential evolution (DE), Evolutionary Multi-objective optimization (EMO) and Memetic computing (MC).

The aim is to investigate in both the new theories and methods on how to transfer learning can be achieved with different evolutionary computation paradigms, and how to 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 evolutionary computation for classification
  • Transfer learning in evolutionary computation for regression
  • Transfer learning in evolutionary computation for clustering
  • Transfer learning in evolutionary computation for other data mining tasks, such as association rules and link analysis
  • Transfer learning in evolutionary computation for numeric optimisation tasks
  • Transfer learning in evolutionary computation for scheduling and combinatorial optimisation tasks
  • Hybridization of evolutionary computation and neural networks, and fuzzy systems and memetic computing for transfer learning
  • Hybridization of evolutionary computation and machine learning, information theory, statistics, etc., for transfer learning
  • Transfer learning in evolutionary computation for real-world applications, e.g. text mining, image analysis, face recognition, WiFi localization, etc.

CEC-9  Special Session on Evolutionary Computer Vision

Organized by Mengjie Zhang (mengjie.zhang@ecs.vuw.ac.nz), Vic Ciesielski, Mario Köppen

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 artificial bee colony optimisation;
  • Emergent, metaheuristics and other EC approaches such as learning classifier systems, differential evolution, artificial immune systems, multi-objective optimisation, hybrid search and memetic computing, transfer learning and domain adaptation, deep learning and kernel methods, hyper-heuristic techniques; and
  • Cross-fertilization of evolutionary computation with other techniques such as neural networks including deep learning and fuzzy systems is also encouraged. New methods in evolutionary deep learning and evolutionary fuzzy deep learning for image and vision computing are highly welcome.

Applications in computer vision and image processing including

  • Edge 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-10  Special Session on Evolutionary Computation for Service-Oriented Computing

Organized by Hui Ma, Yi Mei (yi.mei@ecs.vuw.ac.nz), and Mengjie Zhang

Service-oriented computing is becoming more and more prominent in the Internet environment with the rapid growth of Web services available on the internet. This raises issues for Web service providers such as Web service composition and location allocation, resource allocation and scheduling, etc. Furthermore, there are multiple potentially conflicting objectives (called Quality-of-Service, QoS) to be considered simultaneously in the problem such as response time, cost, reliability, safety, etc. In the era of cloud computing and Big Data, the number and complexity of Web services on the Internet is increasing rapidly. Traditional service composition approaches have come to a performance bottleneck.

Computational Intelligence (CI) has been successfully applied to many challenging real-world problems. This special session aims to solve the service-oriented computing problems with CI techniques, covering all different evolutionary computation paradigms such as Genetic Algorithms (GAs), Genetic Programming (GP), Evolutionary Programming (EP), Evolution Strategies (ES), Memetic Algorithms (MAs), Learning Classifier Systems (LCS), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Differential Evolution (DE), and Evolutionary Multi-objective Optimization (EMO), as well as neural networks and fuzzy systems.

Scope and Topics

The scope of this special session includes both new theories and methods on how to solve the challenging service-oriented computing problems such as Web service composition and location-allocation more effectively and efficiently. Authors are invited to submit their original and unpublished work to this special session.

Topics of interest include, but not limited to:

  • Evolutionary Web service composition
  • Evolutionary Web service workflow optimisation
  • Evolutionary Web service selection
  • Evolutionary Web service location allocation
  • Evolutionary Web service scheduling
  • Evolutionary semantic Web service composition
  • Evolutionary dynamic Web service composition
  • Multi-objective Web service composition
  • Evolutionary computation for resource allocation in Cloud computing
  • Evolutionary computation for workflow management in Cloud
  • Evolutionary computation for distributed Web service composition
  • Novel representations and search operators for Service-oriented computing
  • Cooperative coevolution for Service-oriented computing
  • Neural networks for service-oriented computing
  • Fuzzy systems for service-oriented computing
  • Hybrid algorithms between EC, neural networks and fuzzy systems for service-oriented computing

CEC-11  Special Session on Advanced Evolutionary Computation Approaches for Smart Grid and Sustainable Energy Systems

Organized by: Zhi-le Yang (zl.yang@siat.ac.cn), Jing J. Liang and Kang Li

To shape a low carbon energy future has been a crucial and urgent task under Paris Global Agreement. Numerous optimisation problems have been formulated and solved to effectively save the fossil fuel cost and relief energy waste. However, the majority of the problems are of strong non-convex and non-smooth characteristics. Evolutionary computation is promising to provide powerful optimisation tools for intelligently and efficiently solving problems such as power and sustainable energy systems scheduling to reduce carbon consumptions.

Scope and Topics

This special session intends to bring together the state-of-the-art advances in evolutionary optimisation approaches for solving emerging problems in complex modern power and sustainable energy system. The submissions are encouraged to be the focus on smart grid scheduling with the integration of new participants such as renewable generations, plug-in electric vehicles, distribution generations and energy storages, multiple time-spacial energy reductions and other energy optimisation topics.

Potential submission topics include not but not limited to the below ones:

  • Unit commitment, economic dispatch, and optimal power flow
  • Optimal smart grid scheduling and integration with renewable energy generations
  • Intelligent coordination and control of plug-in electric vehicles
  • Efficient powertrain management for hybrid electric vehicles
  • Charging and discharging strategies for energy storage battery systems
  • Internal and whole scale management for single and hybrid energy storage systems
  • Energy reduction strategies for energy-intensive manufacturing processes
  • Parameters identification for photovoltaic models and PEM fuel cells
  • Thermodynamic optimisation for heat exchanger design and Organic Rankine Cycle