A Dynamic Density Peak Clustering Algorithm Based on K-Nearest Neighbor

The clustering results of the density peak algorithm (DPC) are greatly affected by parameter d c , and center needs to be selected manually. To solve these problems, this paper proposes a low sensitivity dynamic based on K-Nearest Neighbor (DDPC), label is allocated adaptively analyzing distribution Neighbors around each data. It reduces eliminates selecting centers manually from decision graph. Through experimental analysis comparison artificial dataset UCI dataset, show that comprehensive effect DDPC better than DPC, DBSCAN, DBC, other algorithms.