Discrete differential evolution based on estimation of distribution

Differential evolution(DE) is the latest intelligent algorithms for solving the optimization problems very effectively.The algorithm is mainly used in solving the global continuous optimization,but their applications to combinatorial optimization have been rather limited and are not as effective as in global continuous optimization.Firstly a discrete differential evolution(DE) for combinatorial optimization is proposed,and then incorporates the estimation of distribution algorithm(EDA) into the discrete DE to improve its performance.The proposed discrete DE algorithm based on EDA combine global statistical information extracted by EDA with local evolution information obtained by discrete DE to create promising solutions.In order to keep the diversities in the population,a bit flip mutation operator is also incorporated into the proposed hybrid algorithm.The results of experiment show that the EDA can significantly improve the performance of the discrete DE.