Unsupervised dimensionality reduction by jointing dynamic hypergraph and low-rank embedding for classification and clustering

作者：

Highlights：

• An unsupervised dimensionality reduction method called DHLRE is proposed.

• DHLRE is found to be able to extract features with robustness and discrimination.

• A new hypergraph construction method without neighborhood parameter is proposed.

• An efficient optimization algorithm is developed to solve DHLRE model.

摘要

•An unsupervised dimensionality reduction method called DHLRE is proposed.•DHLRE is found to be able to extract features with robustness and discrimination.•A new hypergraph construction method without neighborhood parameter is proposed.•An efficient optimization algorithm is developed to solve DHLRE model.

论文关键词：Unsupervised dimensionality reduction,Dynamic hypergraph learning,Low-rank representation,Classification,Clustering

论文评审过程：Received 10 March 2022, Revised 6 July 2022, Accepted 17 July 2022, Available online 20 July 2022, Version of Record 26 July 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118225