Experimental Supplements to the Theoretical Analysis of Migration in the Island Model
In Lässig and Sudholt (GECCO 2010) the first running time analysis of a non-trivial parallel evolutionary algorithm was presented. It was demonstrated for a constructed function that an island model with migration can drastically outperform both panmictic EAs as well as parallel EAs without migration. This work provides additional empirical results that increase our understanding of why and when migration is essential for this function. We provide empirical evidence complementing the theoretical results, investigate the robustness with respect to the choice of the migration interval and compare various migration topologies using statistical tests.
KeywordsGlobal Optimum Success Probability Good Individual Experimental Supplement Island Model
Unable to display preview. Download preview PDF.
- 2.He, J., Yao, X.: Analysis of scalable parallel evolutionary algorithms. In: Proceedings of CEC 2006, pp. 120–127 (2006)Google Scholar
- 5.Skolicki, Z., De Jong, K.: The influence of migration sizes and intervals on island models. In: Proceedings of GECCO 2005, pp. 1295–1302. ACM, New York (2005)Google Scholar
- 6.Lässig, J., Sudholt, D.: The benefit of migration in parallel evolutionary algorithms. In: Proceedings of GECCO 2010, pp. 1105–1112. ACM, New York (2010)Google Scholar
- 7.Wineberg, M., Christensen, S.: Statistical analysis for evolutionary computation: introduction. In: Companion of GECCO 2009, pp. 2949–2976. ACM, New York (2009)Google Scholar