Fuzzy c-means clustering-based mating restriction for multiobjective optimization

Mating restriction is an important approach to improve the performance of multiobjective evolutionary algorithms (MOEAs). This paper designs a fuzzy c-means clustering-based mating restriction (FMR) scheme, and proposes a fuzzy c-means clustering-based MOEA named as FCMMO. FMR employs a fuzzy c-means clustering algorithm to discover the clustering structure of the solutions in the population. Based on the structure, a specific mating pool is determined for each solution to generate new solutions. Experimental studies show that FCMMO outperforms five state of the art MOEAs on a set of test instances with complicated Pareto front shapes and Pareto set structures, and FMR significantly contributes to the good performance of FCMMO.