KR-DBSCAN: A density-based clustering algorithm based on reverse nearest neighbor and influence space

作者：

Highlights：

• We define a new cluster expanding condition for the density-based clustering.

• We design a new noise removal approach in the density-based clustering analysis.

• We propose a density-based clustering algorithm KR-DBSCAN.

• KR-DBSCAN is evaluated through extensive experiments.

摘要

•We define a new cluster expanding condition for the density-based clustering.•We design a new noise removal approach in the density-based clustering analysis.•We propose a density-based clustering algorithm KR-DBSCAN.•KR-DBSCAN is evaluated through extensive experiments.

论文关键词：Density-based clustering,Cluster expansion,Reverse nearest neighborhood,Influence space,Core object

论文评审过程：Received 5 June 2020, Revised 20 March 2021, Accepted 11 August 2021, Available online 25 August 2021, Version of Record 31 August 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115763