Abstract

The paper deals with the Bivariate Marginal Distribution Algorithm (BMDA). BMDA is an extension of the Univariate Marginal Distribution Algorithm (UMDA). It uses the pair gene dependencies in order to improve algorithms that use simple univariate marginal distributions. BMDA is a special case of the Factorization Distribution Algorithm, but without any problem specic knowledge in the initial stage. The dependencies are being discovered during the optimization process itself. In this paper BMDA is described in detail. BMDA is compared to dierent algorithms including the simple genetic algorithm with dierent crossover methods and UMDA. For some tness functions the relation between problem size and the number of tness evaluations until convergence is shown. 1. Introduction Genetic algorithms work with populations of strings of xed length. In this paper binary strings will be considered. From current population better strings are selected at the expense of worse ones. New strings ar...

Citations