General density-peaks-clustering algorithm
Tianyi Li, Shihong Yue, Chang Ku Sun · 2021
Density-peaks-clustering (DPC) algorithm plays an important role in clustering analysis with the advantages of easy realization and comprehensiveness whereas without the requirement of any iteration or optimization. However, the DPC accuracy depends on two user-specified parameters, and each of them can greatly affect clustering results. To solve this problem, we extend DPC to a general and hierarchical form. Without the need of any parameters, the proposed E-DPC algorithm can effectively cluster points in any dataset with various characteristics. The results of experiments show that the proposed algorithm is more accurate and general in comparison with two mostly used algorithms.