An intelligent swarm clustering algorithm using swarm similarity measure

Abstract

Using big data analysis technologies to cluster user data and mine the potential of stock users is of great significance to telecom operators. In this paper, we present an intelligent swarm clustering algorithm using swarm similarity measure for this purpose. In the proposed scheme, according to the definition of cluster center and the probability distribution of the similarity between objects and cluster centers, we accomplish the cluster process. The improved algorithm can meet requirements of customer clustering and choose clustering centers without human supervision. Moreover, the algorithm can adjust the similarity and its probability distribution decided by different features, so it can work more flexibly and efficiently.