Benchmarking a $(\mu+\lambda)$ Genetic Algorithm with Configurable Crossover Probability

TL;DR

This paper empirically investigates how varying crossover probability in $(+)$ GAs affects performance across different optimization problems, revealing nuanced insights into the interplay of crossover, mutation, and population size.

Contribution

It introduces a configurable crossover probability in $(+)$ GAs and provides empirical analysis of its impact on diverse optimization tasks, highlighting non-asymptotic effects.

Findings

Crossover-based configurations perform better on easy problems.

Fast mutation with crossover outperforms standard mutation on complex tasks.

Optimal crossover probability varies with population size and problem dimension.

Abstract

We investigate a family of $(\mu+\lambda)$ Genetic Algorithms (GAs) which creates offspring either from mutation or by recombining two randomly chosen parents. By scaling the crossover probability, we can thus interpolate from a fully mutation-only algorithm towards a fully crossover-based GA. We analyze, by empirical means, how the performance depends on the interplay of population size and the crossover probability. Our comparison on 25 pseudo-Boolean optimization problems reveals an advantage of crossover-based configurations on several easy optimization tasks, whereas the picture for more complex optimization problems is rather mixed. Moreover, we observe that the ``fast'' mutation scheme with its are power-law distributed mutation strengths outperforms standard bit mutation on complex optimization tasks when it is combined with crossover, but performs worse in the absence of crossover. We then take a closer look at the surprisingly good performance of the crossover-based $(\mu+\lambda)$ GAs on the well-known LeadingOnes benchmark problem. We observe that the optimal crossover probability increases with increasing population size $\mu$. At the same time, it decreases with increasing problem dimension, indicating that the advantages of the crossover are not visible in the asymptotic view classically applied in runtime analysis. We therefore argue that a mathematical investigation for fixed dimensions might help us observe effects which are not visible when focusing exclusively on asymptotic performance bounds.