A New Approach of Data Clustering Using Quantum Inspired Particle Swarm Optimization Based Fuzzy c-means

Abstract

In this article, a Quantum inspired Particle swarm optimization (QtPSO) based Fuzzy c-means algorithm is proposed to cluster multidimensional data. Sometimes, fuzzy c-means used to get stuck at local minima due to the improper selection of cluster centers initially. To sort out the drawback, the intended QtPSO algorithm is applied to generate the cluster centers for a dataset. The effectivity of quantum computing is melted with the well known PSO algorithm. For designing this proposed algorithm, the feature of qubit is applied in association with particle swarm optimization. The proposed algorithm has been compared rigorously with the conventional fuzzy c-means algorithm and modified quantum-inspired particle swarm optimization algorithm (MQPSO) on four well known dataset. The superiority of the proposed algorithm is demonstrated on the basis of two standard cluster evaluation criteria, min value, max value, mean Value, median Value, standard deviation and best convergence times, mean convergence times and one statistical significance test, called Kruskal-Wallis H-test for different levels of clustering.