An Efficient Density-based Clustering Algorithm Combined with Representative Set
Hongfang Zhou · Journal of Information and Computational Science · 2013
Cluster analysis is a key technique of data mining and has been applied widely. The paper presents a novel clustering algorithm DCURS, which incorporates the principle of proximity and limited area into the density-based clustering method to improve clustering accuracy and stability. In addition, DCURS introduces the representative set to accelerate the running speed of the algorithm. Experimental results verify that DCURS has better performance than traditional density-based algorithms of DBSCAN and DBRS.