Assortative mating in genetic algorithms for dynamic problems

Citation

Ochoa G, Madler-Kron C, Rodriguez R & Jaffe K (2005) Assortative mating in genetic algorithms for dynamic problems. In: Rothlauf F, Branke J, Cagnoni S, Wolfe CD, Dreschsler R, Jin Y, Machado P, Marchiori E, Romero J, Smith G & Squillero G (eds.) Applications of Evolutionary Computing: EvoWorkkshops 2005: EvoBIO, EvoCOMNET, EvoHOT, EvoIASP, EvoMUSART, and EvoSTOC Lausanne, Switzerland, March 30 - April 1, 2005 Proceedings. Lecture Notes in Computer Science, 3449. EvoWorkshops 2005: EvoBIO, EvoCOMNET, EvoHOT, EvoIASP, EvoMUSART, and EvoSTOC, Lausanne, Switzerland, 30.03.2005-01.04.2005. Berlin Heidelberg: Springer, pp. 617-622. http://link.springer.com/chapter/10.1007/978-3-540-32003-6_65#

Abstract
Non-random mating seems to be the norm in nature among sexual organisms. A common mating criteria among animals is assortative mating, where individuals mate according to their phenotype similarities (or dissimilarities). This paper explores the effect of including assortative mating in genetic algorithms for dynamic problems. A wide range of mutation rates was explored, since comparative results were found to change drastically for different mutation rates. The strategy for selecting mates was found to interact with the mutation rate value: low mutation rates were the best choice for dissortative mating, medium mutation values for the standard GA, and higher mutation rates for assortative mating. Thus, GA efficiency is related to mate selection strategies in connection with mutation values. For low mutation rates typically used in GA, dissortative mating was shown to be a robust and promising strategy for dynamic problems.