On the Dynamics of EAs without Selection.

This paper investigates the dynamics of evolutionary algorithms (EAs) without tness based selection (constant tness) and finds that such algorithms exhibit a behavior similar to the MISR eeect (mutation-induced speciation by recombination) which has been found in the analysis of (== D ;) evolution strategies.

Abstract

This paper investigates the dynamics of evolutionary algorithms (EAs) without fitness based selection (constant fitness). Such algorithms exhibit a behavior similar to the MISR effect (mutation-induced speciation by recombination) which has been found in the analysis of (=D ; ) evolution strategies. It will be shown that this behavior can be observed in a variety of EAs, not only in unrestricted search spaces, but also in binary GAs. The quantification of this effect is done by introducing the expected population variance oe 2 P . The evolution of oe 2 P over the time g is analytically calculated for both unrestricted and binary search spaces. The theoretical predictions are compared with experiments. The genetic drift phenomenon and the diffusion effect are derived from the general oe 2 P formulae, and it will be shown that MISR is a finite population size sampling effect which cannot be observed in infinite populations. 1 Introduction The question of how GAs, or more generally...