Estimation of Distribution Algorithm for Mixed Continuous-Discrete Optimization

In recent few years expressive progress in the theory and practice of Estimation of Distribution Algorithms (EDA) [1] has appeared, where the classical genetic recombination operators are replaced by probability estimation and stochastic sampling techniques. In this paper we identify some disadvantages of present probabilistic models used in EDAs and propose more general and efficient model for continuous optimization problems based on the decision trees. The new variant of EDA is capable to solve mixed continuous-discrete optimization problems.