Influence of Migration on Efficacy and Efficiency of Parallel Evolutionary Computing

Metaheuristics, such as evolutionary algorithms (EAs), have been proven to be (also theoretically, see, for example, the works of Michael Vose [1]) universal optimization methods. Previous works (Zbigniew Skolicki and Kenneth De Jong [2]) investigated impact of migration intervals on island models of EAs in their works. Here we explore different migration intervals and amounts of migrating individuals, complementing Skolicki and DeJong’s research. In our experiments, we use different ways of selecting migrants and pave the way for further research, e.g., involving different topologies and neighborhoods. We present the idea of the algorithm, show experimental results.