Clustering of Fuzzy Data Sets Based on Particle Swarm Optimization with Fuzzy Cluster Centers

Abstract

In the current study, a particle swarm clustering method is suggested for clustering of triangular fuzzy data. This clustering method can find the centers of the fuzzy cluster. Whatever the centers of the fuzzy cluster have more points from the corresponding cluster, clustering accuracy increases. Triangular fuzzy numbers are utilized to demonstrate uncertain data. To compare triangular fuzzy numbers, a similarity criterion based on the intersection region of the fuzzy numbers is used. The performance of the suggested clustering method has experimented on both benchmark and artificial datasets. These datasets are used in the fuzzy form. The experiential results represent that the suggested clustering method with fuzzy cluster centers can cluster triangular fuzzy datasets like or superior to other standard uncertain data clustering methods. Experimental results demonstrate that, in almost all datasets, the proposed clustering method provides better results in accuracy when compared to Uncertain K-Means and Uncertain K-medoids algorithms.