A Game-Theoretic Approach for Designing Mixed Mutation Strategies
Different mutation operators have been proposed in evolutionary programming. However, each operator may be efficient in solving a subset of problems, but will fail in another one. Through a mixture of various mutation operators, it is possible to integrate their advantages together. This paper presents a game-theoretic approach for designing evolutionary programming with a mixed mutation strategy. The approach is applied to design a mixed strategy using Gaussian and Cauchy mutations. The experimental results show the mixed strategy can obtain the same performance as, or even better than the best of pure strategies.
Unable to display preview. Download preview PDF.
- 1.Fogel, D.: Evolution Computation: Toward a New Philosophy of Machine Intelligence. IEEE Press, Piscataway (1995)Google Scholar
- 7.Dutta, P.K.: Strategies and Games. The MIT Press, Cambridge (1999)Google Scholar
- 8.Ficici, S.G., Melnik, O., Pollack, J.B.: A game-theoretic investigation of selection methods used in evolutionary algorithms. In: Proc. of 2000 Congress on Evolutionary Computation, pp. 880–887. IEEE Press, Los Alamitos (2000)Google Scholar
- 9.Wiegand, R.P., Liles, W.C., De Jong, K.A.: Analyzing coperative coevolution with evolutionary game theory. In: Proc. of 2002 Congress on Evolutionary Computation, pp. 1600–1605. IEEE Press, Los Alamitos (2002)Google Scholar