论文信息 - Augmenting SPEA2 with K-Random competitive coevolution for enhanced evolutionary multi-objective optimization

Augmenting SPEA2 with K-Random competitive coevolution for enhanced evolutionary multi-objective optimization

One new algorithm is being proposed, which is the integration between one multiobjective evolutionary algorithm (MOEA): strength Pareto evolutionary algorithm 2 (SPEA2) and competitive coevolution using k-random opponents competitive fitness strategy. The resulting algorithm is referred to as SPEA2-CE-KR. This proposed algorithm was benchmarked against the original SPEA2 using seven DTLZ test problems having 3 to 5 objectives. Overall, reveal that SPEA2-CE-KR performed well for the spacing and coverage metrics.

Hui Keng Lau | Tse Guan Tan | J. Teo

[1] Lothar Thiele,et al. Comparison of Multiobjective Evolutionary Algorithms: Empirical Results , 2000, Evolutionary Computation.

[2] Richard K. Belew,et al. Methods for Competitive Co-Evolution: Finding Opponents Worth Beating , 1995, ICGA.

[3] Lothar Thiele,et al. A Tutorial on the Performance Assessment of Stochastic Multiobjective Optimizers , 2006 .

[4] Jason R. Schott. Fault Tolerant Design Using Single and Multicriteria Genetic Algorithm Optimization. , 1995 .

[5] Carlos A. Coello Coello,et al. Evolutionary Multi-Objective Optimization: A Critical Review , 2003 .

[6] Liviu Panait. A Comparative Study of Two Competitive Fitness Functions , 2002 .

[7] Robert Axelrod,et al. The Evolution of Strategies in the Iterated Prisoner's Dilemma , 2001 .

[8] Marco Laumanns,et al. Scalable Test Problems for Evolutionary Multiobjective Optimization , 2005, Evolutionary Multiobjective Optimization.

[9] Marco Laumanns,et al. SPEA2: Improving the strength pareto evolutionary algorithm , 2001 .

[10] Peter J. Angeline,et al. Competitive Environments Evolve Better Solutions for Complex Tasks , 1993, ICGA.