A new density based clustering algorithm for Binary Data sets

Binary Data clustering finds tremendous applications in fault analysis of machineries, document classification, image retrievals and analysis, medical diagnosis of diseases etc. Accurate clustering of binary databases provides numerous help to develop accurate designs of above systems. In this manuscript a density based clustering algorithm is proposed to effectively cluster binary datasets. The proposed algorithm automatically determines the number of clusters based upon the density of data present in a region. The number of clusters evolve during the clustering process due to merging of several smaller clusters. Simulation studies were carried out on two synthetic datasets and it is observed that the proposed algorithm can effectively clusters both correlated and random binary datasets.