Improving fuzzy c-means clustering viaquantum-enhanced weighted superpositionattraction algorithm

Abstract

Fuzzy clustering has become an important research field in patternrecognition and data analysis. As supporting unsupervised mode oflearning, fuzzy clustering brings about unique opportunities to revealstructural relationships in data. Fuzzy c-means clustering is one ofthe widely preferred clustering algorithms in the literature. However,fuzzy c-means clustering algorithm has a major drawback that it canget trapped at some local optima. In order to overcome this short-coming, this study employs a new generation metaheuristic algorithm.Weighted Superposition Attraction Algorithm (WSA) is a novel swarmintelligence-based method that draws inspiration from the superposi-tion principle of physics in combination with the attracted movementof agents. Due to its high converging capability and practicality, WSAalgorithm has been employed in order to enhance performance of fuzzy-c means clustering. Comprehensive experimental study has been con-ducted on publicly available datasets obtained from UCI machine learn-ing repository. The results point out significant improvements over thetraditional fuzzy c-means algorithm