Semi-supervised discriminative clustering with graph regularization

作者：

Highlights：

• We introduce discriminative semi-supervised clustering model.

• It applies pairwise constraints as side information.

• Our model maximizes the number of correctly classified pairs of data points.

• It outperforms existing methods in the experimental study.

摘要

•We introduce discriminative semi-supervised clustering model.•It applies pairwise constraints as side information.•Our model maximizes the number of correctly classified pairs of data points.•It outperforms existing methods in the experimental study.

论文关键词：Semi-supervised clustering,Discriminative model,Pairwise constraints,Graph clustering

论文评审过程：Received 23 October 2017, Revised 2 March 2018, Accepted 11 March 2018, Available online 12 March 2018, Version of Record 11 May 2018.

论文官网地址：https://doi.org/10.1016/j.knosys.2018.03.019