Comparative study of density-based clustering algorithms

Abstract

Clustering is an unsupervised learning. It will divide clusters without assigning labels. The process of partitioning the data into groups known as clusters in such a way that the intraclass similarity is high and interclass similarity is low. This paper is proposed to give a comparative study of various density-based clustering algorithms of data mining. The following are different density-based clustering algorithms which will be reviewed in this paper: DBSCAN, OPTICS, and DENCLUE.