Multi-Chromosomal Representations in Neuroevolution

Contemplating the development of the field of evolutionary computation (EC), where most basic concepts are borrowed from nature, it is remarkable that multi-chromosomal representations present in all complex organisms have rarely been studied in the artificial domain. Evidently, the addition of such an additional layer of genetic code must prove to possess certain benefits before it is introduced. Thus, we present experiments with multi-chromosomal evolution of artificial neural networks (ANNs) on three benchmark problems with the goal to investigate potential advantages of genotypes with multiple chromosomes. Besides the technical benefit of using different encodings, genetic operators, and parameters for specific chromosomes, we hypothesize that the generalization capabilities of evolved networks (or other structures) increase, when their genotype is of multi-chromosomal organization.
Helmut A. Mayer

Last modified: Dec 22 2006