Home > Research
Workshop on Selection
AbstractA recurring research theme in metaheuristics is to consider the balance between Exploration and Exploitation. An often forgotten area of research is the balance between Selection and Exploration. It is noted that Selection has the ability to turn any exploratory search process into a hill climber by rejecting all exploratory search solutions. Recent experiments show that this occurs in Particle Swarm Optimization where the rejection of all exploratory particle positions can cause the swarm to stall before it can converge. The first goal of this workshop will be to reanalyze current metaheuristics from the perspective of selection (as opposed to exploration and/or an underlying metaphor). Subsequent results/goals include a selection-based taxonomy for the explosion of metaphor-based metaheuristics, tools to accurately measure exploration and the effects of selection on exploratory search solutions, the identification and categorization of selection errors, and suggestions for future methods of selection and metaheuristic design.
Workshop Overview
Full details on our Research Gate site (see Project log)
Workshop Papers
S. Chen, S. Islam, and S. Lao. (2021) "Simulated Annealing vs. Hill Climbing in Continuous Domain Search Spaces." Workshop on Selection at CEC2021. Kraków, Poland (VIRTUAL). June 2021.
J. Montgomery, A. Bolufé Röhler, D. Tamayo-Vera, S. Chen, and T. Hendtlass. (2021) "Some Convergence Properties of Differential Evolution." Workshop on Selection at CEC2021. Kraków, Poland (VIRTUAL). June 2021.
N. Yadollahpour and S. Chen. (2021) "Stall Detection in Particle Swarm Optimization." Workshop on Selection at CEC2021. Kraków, Poland (VIRTUAL). June 2021.
I. Moser. (2021) "Algorithm Footprinting for Search Behaviour Characterisation." Workshop on Selection at CEC2021. Kraków, Poland (VIRTUAL). June 2021.
A. Bolufé Röhler and S. Chen. (2021) "Observations on Convergence in Leaders and Followers." Workshop on Selection at CEC2021. Kraków, Poland (VIRTUAL). June 2021.
Program
Our Program. slot is Monday, June 28 in Room 7 from 14:15-16:15 Central European Summer Time.
14:15 - 14:20 -- Introductions
Opening remarks
14:20 - 14:40 -- Swarm Intelligence
Workshop Paper: Stall Detection in Particle Swarm Optimization (video)
Naeemeh Yadollahpour
Commentary: Artificial Bee Colony
Mohammed El-Abd
Open Discussion
14:40 - 15:00 -- Evolutionary Computation
Workshop Paper: Some Convergence Properties of Differential Evolution (video)
James Montgomery
Commentary: Grey Wolf Optimizer
Seyed Mohammad Mirjalili
Open Discussion
15:00 - 15:40 -- Search Space Considerations
Workshop Paper: Algorithm Footprinting for Search Behaviour Characterisation (video)
Irene Moser
Invited Presentation: Challenges in computing attraction basins for continuous functions (video)
Marjan Mernik, Mihael Baketaric, Jernej Jerebic, Shih-Hsi Liu, Miha Ravber, Luka Mernik, Matej Crepinšek
Invited Presentation: If you can’t beat it, squash it! Simplify global optimization by evolving dilation functions (video)
Marco Nobile
Open Discussion
15:40 - 16:05 -- Failures of Selection
Workshop Paper: Observations on Convergence in Leaders and Followers (video)
Antonio Bolufé-Röhler
Workshop Paper: Simulated Annealing vs. Hill Climbing in Continuous Domain Search Spaces (video)
Stephen Chen
Open Discussion
16:05 - 16:15 -- Closing Remarks
Open Discussion
Workshop Organizers
Stephen Chen is an Associate Professor in the School of Information Technology at York University in Toronto, Canada. His research focuses on analyzing the mechanisms for selection, exploration, and exploitation in techniques designed for multi-modal optimization problems. He is particularly interested in the development and analysis of non-metaphor-based heuristic search techniques. He has conducted extensive research on Genetic algorithms and Particle Swarm Optimization. He has over 60 peer-reviewed publications including 26 CEC papers. He has also presented 3 tutorials at CEC/WCCI events and is a Senior Member of the IEEE.
Tim Hendtlass is an Adjunct Professor in the School of Software and Electrical Engineering at Swinburne University. He has worked in artificial intelligence for some 35 years, has about 80 refereed publications and regularly reviews for eight different journals in the field, including three IEEE Transactions. To date he has conducted over 150 reviews. His personal research interests lie with evolutionary processes and particle swarm optimisation, as well as artificial neural networks. His most recent works investigate the advantage to be gained by separating exploratory and exploitative search mechanism in population-based search.
Irene Moser is an Associate Professor in the Department of Computer Science and Software Engineering at Swinburne University of Technology in Melbourne, Australia. One of her research interests is the characterisation of the search process of evolutionary algorithms and other stochastic heuristics with a view to estimating the quality of the results they achieve. She takes a keen interest in practical applications of optimisation algorithms in areas like urban design, sustainability, transport and data science, and the development of custom optimisation algorithms for particular problems. She has over 80 publications in peer-reviewed venues, including 10 CEC papers. She has also presented a tutorial at GECCO and is a Senior Member of the IEEE.
James Montgomery is a Senior Lecturer in the School of Information and Communication Technology at the University of Tasmania, based in Hobart, Australia. His evolutionary computation research spans both the analysis of search algorithm behaviour and the application of EC techniques to real-world, multiobjective problems, including sustainable agricultural planning in a changing climate. Outside of EC he works in the growing field of ecoacoustics, the application of machine learning to environmental audio recordings to derive ecological insights. He has deep knowledge of Ant Colony Optimization and Differential Evolution. He has published over 65 peer-reviewed papers (more than 40 of which are on EC, 17 of which were presented at CEC) and has delivered three CEC/WCCI tutorials.
Antonio Bolufé Röhler is an Assistant Professor at the University of Prince Edward Island. He obtained a Bachelor’s in Computer Science and a Master’s and PhD in Mathematics from the University of Havana. His main research focuses on understanding and formalizing (meta)heuristic optimization. He is particularly interested in the design of new non-metaphor-based and hybrid heuristics, large scale optimization and the use of machine learning for improving heuristic search. Relevant contributions include research on Minimum Population Search, Leaders and Followers, and Exploration-only Hybrids. To date he has over 30 peer-reviewed publications, has conducted over 50 reviews, and is Executive Secretary of the Gecontec journal.
Mohammed El-Abd (SM'16) is an Associate Professor of Computer Engineering in the College of Engineering and Applied Sciences at The American University of Kuwait (AUK). He obtained his B.Eng. and M.Sc. from the ECE Department at Ain Shams University in Egypt in 1998 and 2003, respectively. He obtained his Ph.D. from the ECE Department at the University of Waterloo (UW) in Canada in 2008. He has published more than 10 journal articles and over 60 conference publications, abstracts, and book chapters. He is an Associate Editor of the Swarm and Evolutionary Computation (SWEVO) journal by Elsevier since 2016 and served as a Guest Editor for the IEEE Transactions on Education. He serves as a reviewer for many prestigious journals including the IEEE Transactions on Evolutionary Computation, the IEEE Transactions on Cybernetics and IEEE Access. Moreover, he was the main organizer of a special session on “Cooperative Evolutionary Computation” in IEEE-CEC in 2019. He is the founding chair of the “Symposium on Cooperative Metaheuristics” organized within IEEE-SSCI. His research interests span the areas of meta-heuristics, evolutionary computation, swarm intelligence, cooperative algorithms, continuous optimization, large-scale optimization, Internet-of-Things (IoT), smart cities, control & robotics, and engineering education.
Seyed Mohammad Mirjalili is a research assistant with the Department of Electrical & Computer Engineering at Concordia University in Montreal, Canada. Seyed Mohammad is internationally recognized for his contributions to the fields of Evolutionary Algorithms and Photonics. He has published 30 journal articles with over 7800 citations in total, with an H-index of 17 from Google Scholar Metrics. He has been one of the inventors of a number of new AI evolutionary algorithms such as Grey Wolf Optimizer and Salp Swarm Algorithm that have been widely used and highly cited across the globe in both academia and industry.
