Bayesian Based Metaheuristic for Large Scale Continuous Optimization

This paper is dedicated to design an efficient met heuristic based on Bayesian ape proach to solve continuous optimization problems. The proposed approach is based on the use of different search strategies (crossover and mutation) and then selects the best strategy from those possible ones based on the Bayes theorem. The obtained results were compared to those obtained using a met heuristic that uses a static strategy in order to show the benefit of changing the search exploration dynamically along the generations. Moreover, we compared the performance of our approach on the CEC 2008 benchmark. These results show its efficiency.