Recent studies of spatially distributed EAs have formallycharacterized the selection pressure induced by various selectionstrategies applied to local neighborhoods of various sizes and shapes.These analyses provide us with the ability to predict the expectedbehavior of the local neighborhood EAs. In this paper we empiricallyvalidate these predictions using the domain of function optimization. Wedemonstrate the various ways selection pressure can be varied in aspatially distributed EA and show that, from a performance point ofview, no optimal selection pressure can be defined since it also dependson the fitness landscape of the problem being solved. Our resultssuggest that it may be possible to adaptively tune selection pressure byvarying a single parameter, the neighborhood radius
展开▼
