.png){kind=icon}
.png){kind=icon}
Return
Multi-view low-rank matrix factorization using multiple manifold regularization
.png){kind=icon}
PRE
AI
DOI：10.1016/j.neucom.2019.01.004.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Low-rank matrix factorization has been applied to solving many computer vision and machine learning problems. However, most of the existing low-rank matrix factorization methods are usually suitable to handle the single-view data. For the data containing multi-view features, the existing methods ignore the complementary information of multi-view data, resulting in the degenerated learning performance. To effectively learn the suitable representation for multi-view data, we propose a novel multi-view low-rank factorization method. Specifically, a multi-manifold regularizer is adopted to force the consensus data representation to move smoothly on the underlying manifold, where the constructed hypergraphs with the suitable weights are combined to estimate the manifold of multi-view data. Furthermore, the multi-manifold regularizer is incorporated into the multi-view low-rank matrix factorization framework to simultaneously learn the hypergraph weights and the consensus representation matrix. The learned compact representation of low-rank matrix factorization effectively discovers the underlying consistent semantics across different view features, which achieves promising clustering performance on real-world datasets. (C) 2019 Elsevier B.V. All rights reserved.
Keywords:
Hypergraph learning
Low-rank
Multi-view
Sparse representation
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
6.5
Papers:
2.5W
Citations:
6.5W