Density peak clustering evaluation

Density based clustering is one of the most tools used in clustering data analysis because of it can detect cluster with arbitrary shapes and noises. Some methods have been proposed in literature such as DBSCAN and density peak clustering. In density peak clustering (DPC), the results may directly depend on the method used in density estimation. In this paper, we conduct some experimentations with various data set to observe and analysis the results obtained by DPC when using some kinds of density estimation.