Learn reversi using parallel genetic algorithms

Abstract

This paper presents how parallel genetic algorithms can be used to learn the game Reversi. Parallel Genetic Algorithm can create a distributed environment in which instances of genetic algorithms are executed in parallel. The history of chromosomes is stored in an ontology using OWL language. One computer player takes a decision using the Analytic Hierarchy Process (AHP) method of multiple criteria decision analyses (MCDA) and using game decision tree. To learn the game means to find the optimal weights and rules. © 2008 Springer-Verlag Berlin Heidelberg.