A robust clustering algorithm based on the identification of core points and KNN kernel density estimation

作者：

Highlights：

• A new method is proposed to eliminate the drawbacks of the original DPC algorithm.

• A new form of KNN kernel function is applied to estimate local density distribution.

• Clusters are represented by core points which reveal the structure of clusters.

• The new method performs better than widely used clustering algorithms.

摘要

•A new method is proposed to eliminate the drawbacks of the original DPC algorithm.•A new form of KNN kernel function is applied to estimate local density distribution.•Clusters are represented by core points which reveal the structure of clusters.•The new method performs better than widely used clustering algorithms.

论文关键词：Clustering,Density peaks,Core points,Density estimation,KNN kernel

论文评审过程：Received 10 August 2020, Revised 28 December 2021, Accepted 17 January 2022, Available online 3 February 2022, Version of Record 14 February 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116573