Integrating Multi-Objective Genetic Algorithm and Validity Analysis for Locating and Ranking Alternative Clustering

Clustering algorithms in general need the number of clusters as a priori, which is mostly hard for domain experts to estimate. In this paper, we use Niched Pareto k-means Genetic Algorithm (GA) for clustering. After running the multi-objective GA, we get the pareto-optimal front that gives the optimal number of clusters as a solution set. We analyze the clustering results using several cluster validity techniques proposed in the literature, namely Silhoutte, C index, Dunn’s index, DB index, SD index and S-Dbw index. This gives an idea about ranking the optimal number of clusters for each validity index. We demonstrate the applicability and effectiveness of the proposed clustering approach by conducting experiments using two datasets: Iris and the well-known Ruspini dataset. Povzetek: "[Click here and Enter short Abstract in Slovene language]"