A Review on Automatic Clustering Based on Density Metrics

Abstract

Clustering is one of the most popular fields in the domain of data mining. In big data analysis, lots of computational efforts are required for clustering. Traditionally, there are several approaches have been proposed for clustering of data such as K-means. It is the most popular clustering algorithm. However, existing techniques of cluster required ‘k’ parameter in advanced which puts limit on outcomes of clusters where, ‘k’ is the input parameter assign by user or it is the “ideal” number of cluster. In existing RLClu algorithm user have to pre-assign two minimum thresholds of the local density and the minimum density-based distance. Practically, it is difficult to determine the number of clusters in advance. Therefore, efficient technique is required to detect clustering centers and also to address limitations of previous techniques