Evolving better representations through selective genome growth

Publisher Website
Google Scholar
Abstract
— The choice of how to represent the search space for a genetic algorithm (GA) is critical to the GA's per- formance. Representations are usually engineered by hand and fixedfor the duration of the GA run. Here a new method is described in which the degrees of free- dom of the representation — i.e. the genes - are in- creased incrementally. The phenotypic effects of the new genes are randomly drawn from a space of differ- ent functional effects. Only those genes that initially increase fitnessare kept. The genotype-phenotype map that results from this selection during the construction of the genome allows better adaptation. This effect is illustrated with the NK landscape model. The resulting genotype-phenotype maps are much less epistatic than unselected maps would be, having extremely low values of "K" — the number of fitnesscomponents affected by each gene. Moreover, these maps are exquisitely tuned to the specifics of the epistatic fitness function, creat- ing adaptive landscapes that are much smoother than generic NK landscapes with the same genotype-pheno- type maps, with fitnesspeaks many standard deviations higher. Thus a caveat should be made when making ar- guments about the applicability of generic properties of complex systems to evolved systems. This method may help to solve the problem of choice of representations in genetic algorithms.

This publication has 4 references indexed in Scilit: