Super Large Data Sets Clustering by Means Radial Compression

Clustering analysis is an effective technique for exploring data analysis which has been widely applied to varied tasks. Many classical clustering algorithms do good jobs on their prerequisite, but few of them are scalable when applied to Very large data sets (VLDS). In this study, a novel means radial compression clustering method is proposed to deal with the VLDS. First, the concept of means radial compression is defined to describe theoretical model. Next, mean merging is defined and it is proved that the process of mean merging is an efficient method for the implementation of means radial compression. Then, the members will be assigned to the suitable clusters based on the minimum distance between each member and the centers that is found by means radial compression clustering. The experimental results show that means radial compression algorithm can make better solutions compared with the most well known clustering algorithms as K-means clustering, affinity propagation clustering, hierarchical clustering with time complexity of O(n).