A Modified NSGA-II with Silhouette Coefficient and K-means Clustering

Abstract

This article is a proposition for enhancing the genetic algorithm NSGA-II by some form of hybridization. The later explores the K-means clustering algorithm and the Silhouette coefficient features. It implies two specific phases. First, the right number of clusters generated automatically by K-means clustering is verified by Silhouette coefficient according to a number of iterations. Thereafter, NSGA-II is executed, in turn, for a defined number of iterations within the proposed algorithm. Obtained results of the algorithm for some benchmark test functions are used to illustrate the validity of the article proposition.