Adaptive Operator Selection with Bandits for Multiobjective Evolutionary Algorithm Based on Decomposition

Adaptive Operator Selection with Bandits for Multiobjective Evolutionary Algorithm Based on Decomposition

时间：2013年3月25日 地点：九龙湖校区计算机楼313室

报告简介：

Evolutionary Algorithms (EAs) are stochastic optimization algorithms inspired by the Darwinian evolution theory. EAs have already shown their e?ciency on many application domains. However, the performance of EAs is very sensitive to the settings of their intrinsic parameters, and even worse, there are no general guidelines for an efficient setting. The paradigm, referred to as Adaptive Operator Selection (AOS), provides the on-line autonomous control of the operator that should be applied at each instant of the search, i.e., while solving the problem. This paper proposes a bandit based AOS method, Fitness-Rate-Rank-based Multi-Armed Bandit (FRRMAB). In order to track the dynamics of the search process, it uses a sliding window to record the recent fitness improvement rates achieved by the operators, while employing a decaying mechanism to increase the selection probability of the best operator. Not much work has been done on AOS in multiobjective evolutionary computation since it is very difficult to measure the fitness improvements quantitatively in most Pareto dominance based multiobjective evolutionary algorithms. Multiobjective Evolutionary Algorithm based on Decomposition (MOEA/D) decomposes a multiobjective optimization problem into a number of scalar optimization subproblems and optimizes them simultaneously. Thus, it is natural and feasible to use AOS in MOEA/D. We investigate several important issues on using FRRMAB in MOEA/D. Our experimental results demonstrate that FRRMAB is robust and its operator selection is reasonable. Comparison experiments also indicate that FRRMAB can significantly improve the performance of MOEA/D.

报告人简介：

Prof. Sam Kwong received his B.Sc. degree from the State University of New York at Buffalo, Buffalo, NY, M.A.Sc. degree in electrical engineering from the University of Waterloo, Waterloo, ON, Canada, and Ph.D. degree from the Fernuniversit?t Hagen, Hagen, Germany. From 1985 to 1987, he was a Diagnostic Engineer with Control Data Canada, where he designed the diagnostic software to detect the faulty VLSI chips in Cyber 430 machines. He later joined the Bell Northern Research Canada as a Member of Scientific Staff, where he worked on both the DMS-100 voice network and the DPN-100 data network project. In 1990, he entered the academic community at the City University of Hong Kong as a Lecturer in the Department of Electronic Engineering. Currently, he is a Professor in the Department of Computer Science. In 1996, he was responsible for the software design of the first handheld GSM mobile phone consultancy project which was one of the largest consultancy projects at the City University of Hong Kong. He has coauthored three research books, eight book chapters, and over 250 technical papers. His book entitled "Genetic Algorithm for Control and Signal Processing" published by Springer London, was awarded as the bestseller in 1997. Furthermore, he has served as a consultant for several telecommunications companies. Currently, he is the Associate Editor for the IEEE Transactions on Industrial Informatics, the IEEE Transactions on Industrial Electronics, the Journal of Information Science among other reputable journals. From 2000 to 2010, he also served as Associate Editor of the Journal of Real Time Systems. During February 2000, he held the position of Guest Editor of the IEEE transactions on Industrial Electronics. Prof. Kwong has been heavily involved with professional organizations such as IEEE. Currently, Prof. Kwong serves as a Board of Governor Member for the IEEE Systems, Man and Cybernetics (SMCS). He is also the Conference Coordinator for SMCS and a member under Long Range and Planning for the IEEE SMCS society. During his leadership when he was the chairman of IEEE SMC Hong Kong Chapter from 2009-2011, IEEE SMC Hong Kong Chapter was internationally recognized and awarded "Best Chapter Award" in 2011.

计算机网络和信息集成教育部重点实验室（东南大学）