Evolutionary Algorithm Using Marginal Histogram Models in Continuous Domain

Recently, there has been a growing interest in developing evolutionary algorithms based on probabilistic modeling. In this scheme, the offspring population is generated according to the estimated probability density model of the parents instead of using recombination and mutation operators. In this paper, we propose an evolutionary algorithm using a marginal histogram to model the parent population in a continuous domain. We propose two types of marginal histogram models: the fixed-width histogram (FWH) and the fixed-height histogram (FHH). The results showed that both models worked fairly well on test functions with no or weak interactions among variables. Especially, FHH could find the global optimum with very high accuracy effectively and showed good scale-up with the problem size.