Co-Evolutionary Algorithms Based on Mixed Strategy

Co-Evolutionary Algorithms Based on Mixed Strategy

Wei Hou (Harbin Engineering University & Northeast Agricultural University, China), HongBin Dong (Harbin Engineering University, China) and GuiSheng Yin (Harbin Engineering University, China)
Copyright: © 2013 |Pages: 14
DOI: 10.4018/978-1-4666-3625-5.ch006


Inspired by evolutionary game theory, this paper modifies previous mixed strategy framework, adding a new mutation operator and extending to crossover operation, and proposes co-evolutionary algorithms based on mixed crossover and/or mutation strategy. The mixed mutation strategy set consists of Gaussian, Cauchy, Levy, single point and differential mutation operators; the mixed crossover strategy set consists of cuboid, two-points and heuristic crossover operators. The novel algorithms automatically select crossover and/or mutation operators from a given mixed strategy set, and improve the evolutionary performance by dynamically utilizing the most effective operator at different stages of evolution. The proposed algorithms are tested on a set of 21 benchmark problems. The results show that the new mixed strategies perform equally well or better than the best of the previous evolutionary methods for all of the benchmark problems. The proposed MMCGA has shown significant superiority over others.
Chapter Preview

The Framework Of Evolutionary Algorithm Based On Mixed Strategy

In theory mixed strategies (Dutta, 1999; Ficici, Melnik & Pollack, 2000) have some potential advantages over pure strategies (He & Yao, 2005). Individuals are regarded as players in a game. Each individual will choose a crossover or mutation strategy from its strategy set based on a selection probability and generate an offspring by this strategy.

Complete Chapter List

Search this Book: