A New Way to Obtain the Initial Centroid Clusters in Fuzzy C-Means Algorithm

Abstract

Data clustering is an important task in data mining, image processing and other pattern recognition problems. One of the most popular clustering algorithms is the Fuzzy C-Means (FCM). The performance of the FCM is strongly affected by the selection of the initial centroid clusters. Therefore, choosing a good set of initial centroid clusters is very important for the algorithm. However, it is difficult to select a good set of initial centroid clusters randomly. In this paper, we propose a method to obtain the initial centroid clusters in the FCM to accelerate the process of clustering and improve the quality of the clustering.