Parallel Bivariate Marginal Distribution Algorithm with Probability Model Migration

BMDA, Model migration, parallel architectures

This chapter presents a new concept of parallel Bivariate Marginal Distribution Algorithm (BMDA) using the stepping stone communication model with the unidirectional ring topology. The traditional migration of individuals is compared with a newly proposed technique of probability model migration. The idea of the new adaptive BMDA (aBMDA) algorithms is to modify the classic learning of the probability model (applied in the sequential BMDA). In the proposed strategy, the adap-tive learning of the resident probability model is used. The evaluation of pair dependency, using Pearson's chi-square statistics is influenced by the relevant immigrant pair dependency according to the quality of resident and immigrant subpopulation. Experimental results show that the proposed aBMDA significantly outperforms the traditional concept of migration of individuals.