Combining PSO and FCM for Dynamic Fuzzy Clustering Problems

• Published in: Swarm Intelligence Based Optimization pp. 1 - 8
• Main Authors: Kao, Yucheng, Chen, Ming-Hsien, Hsieh, Kai-Ming
• Format: Book Chapter
• Language: English
• Published: Cham Springer International Publishing 01.01.2014
• Series: Lecture Notes in Computer Science
• Subjects: Data Clustering; Fuzzy c-means; Particle Swarm Optimization
• ISBN: 3319129694; 9783319129693
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-319-12970-9_1

This paper proposes a dynamic data clustering algorithm, called PSOFC, in which Particle Swarm Optimization (PSO) is combined with the fuzzy c-means (FCM) clustering method to find the number of clusters and cluster centers concurrently. Fuzzy c-means can be applied to data clustering problems but the number of clusters must be given in advance. This paper tries to overcome this shortcoming. In the evolutionary process of PSOFC, a discrete PSO is used to search for the best number of clusters. With a specified number of cluster, each particle employs FCM to refine cluster centers for data clustering. Thus PSOFC can automatically determine the best number of clusters during the data clustering process. Six datasets were used to evaluate the proposed algorithm. Experimental results demonstrated that PSOFC is an effective algorithm for solving dynamic fuzzy clustering problems.