Island Model genetic Algorithms and Linearly Separable Problems

Parallel Genetic Algorithms have often been reported to yield better performance than Genetic Algorithms which use a single large panmictic population. In the case of the Island Model Genetic Algorithm, it has been informally argued that having multiple subpopulations helps to preserve genetic diversity, since each island can potentially follow a di erent search trajectory through the search space. On the other hand, linearly separable functions have often been used to test Island Model Genetic Algorithms; it is possible that Island Models are particular well suited to separable problems. We look at how Island Models can track multiple search trajectories using the in nite population models of the simple genetic algorithm. We also introduce a simple model for better understanding when Island Model Genetic Algorithms may have an advantage when processing linearly separable problems.