K-Nets: Clustering through nearest neighbors networks

作者：

Highlights：

• A fast, deterministic, exemplar-based algorithm for linear or non-linear clustering.

• Operates with an intuitive resolution parameter or exact number of clusters.

• Depending on data dimensionality can have single or multi-layer architecture.

• Outperforms well established clustering methods.

摘要

•A fast, deterministic, exemplar-based algorithm for linear or non-linear clustering.•Operates with an intuitive resolution parameter or exact number of clusters.•Depending on data dimensionality can have single or multi-layer architecture.•Outperforms well established clustering methods.

论文关键词：

论文评审过程：Received 25 September 2017, Revised 10 October 2018, Accepted 16 November 2018, Available online 17 November 2018, Version of Record 13 December 2018.

论文官网地址：https://doi.org/10.1016/j.patcog.2018.11.010