Back to Search
Start Over
Multi-view clustering with graph regularized optimal transport.
- Source :
-
Information Sciences . Oct2022, Vol. 612, p563-575. 13p. - Publication Year :
- 2022
-
Abstract
- For a multi-view learning task, it is crucial to assign appropriate weights to each view in order to learn complementary and consistent information across different views. In the field of multi-view clustering, most existing methods have been able to handle the weights of different views. However, these algorithms face the problem of unacceptable time complexity when dealing with large-scale datasets, and the learned similarity matrix fails to satisfy the graph regularization. In this paper, we propose an auto-weight learning method called multi-view clustering with graph regularized optimal transport. First, an anchor-based method is employed to overcome the problem of heavy time complexity when processing large-scale datasets, and it is able to automatically learn an appropriate weight for each view. Second, by introducing optimal transport we learn a regularized doubly-stochastic similarity matrix applicable to multi-view clustering tasks. Third, the optimal regularized anchor graph can be classified into specific clusters by adding a rank constraint. Finally, an effective optimization method is designed to optimize the formulated problem. Comprehensive experiments on multiple real-world datasets demonstrate that the proposed algorithm achieves superior performance to other state-of-the-arts algorithms. [ABSTRACT FROM AUTHOR]
- Subjects :
- *ALGORITHMS
Subjects
Details
- Language :
- English
- ISSN :
- 00200255
- Volume :
- 612
- Database :
- Academic Search Index
- Journal :
- Information Sciences
- Publication Type :
- Periodical
- Accession number :
- 160170518
- Full Text :
- https://doi.org/10.1016/j.ins.2022.08.117