Clustering based Adaptive Differential Evolution for Numerical Optimization
Bilal, Millie Pant, Garima Vig · 2020
In this paper, a fuzzy C-means clustering approach is suggested for segregating the initial population of Differential Evolution (DE) on the basis of the membership function. The proposed algorithm called FCADE further incorporates adaptive crossover and mutation strategies into the segregated population. Three variants of FCADE are proposed and are applied to selected CEC2005 benchmark problems. The results, when compared with some of the well-known adaptive algorithms indicates the competence of the strategies proposed in the present study in enhancing the performance of the DE algorithm.