Novel approach to polygamous selection in genetic Algorithms

Abstract

Genetic algorithms are adaptive heuristic search algorithms which have been successfully used in a number of applications and their performance are mainly influenced by selection operator. In this paper three variants of polygamous selection, a special case of elitism where the best individual of the population act as one parent for mating with other chromosomes in all crossover operations, are proposed, their performances are compared along with other selection approaches such as roulette wheel, rank, annealed etc. © 2012 Springer-Verlag GmbH Berlin Heidelberg.