Parallel GEAs with linkage analysis over grid

Abstract

This chapter describes the latest trends in the field of parallel/ distributed computing and the effect of these trends over Genetic and Evolutionary Algorithms (GEAs) especially linkage based GEAs. We concentrate mainly on the Grid computing paradigm which is widely accepted as the most distributed form of computing; due to the advent of Service Oriented Architecture (SOA) and other technologies, Grid has gained a lot of attention in the recent years. We also present a framework that can help users in implementation of metaheuristics based optimization algorithms (including GEAs) over a Grid computing environment. We call this framework MetaHeuristics Grid (MHGrid). Moreover, we give a theoretical analysis of the maximum speed-up achievable by using MHGrid. We also discuss our experience of working with Grids. © 2008 Springer-Verlag Berlin Heidelberg.