QCC:A novel cluster algorithm based on Quasi-Cluster Centers

Abstract

Abstract-Cluster analysis is aimed at classifying elements into categories on the basis of their similarity. And cluster analysis has been widely used in many areas such as pattern recognition, and image processing. In this paper, we propose an approach based on the idea that the density of cluster centers are highest in its k nearest neighborhood or reverse k nearest neighborhood, and clusters is divided by sparse region. We firstly define the similarity between clusters. Based on this idea, no matter non-spherical data or complex manifold data, the proposed algorithm is applicable. And the proposed algorithm has a certain capacity on outliers detection. We demonstrate the power of the proposed algorithm on several test cases. Its clustering performance is better than DBSCAN, DP and K-AP clustering algorithms.