Name
Abstract | ||
|---|---|---|
With the progress of science and technology, multi-view clustering exploring heterogeneous information among diverse views has been widely employed in real-world applications. Nonnegative Matrix Factorization (NMF) has received wide attention because of its interpretability. However, existing multi-view clustering methods based on NMF are vulnerable to outliers and noise. To alleviate these problems, we propose a novel model, named graph regularized multiple nonnegative matrix factorization with L-2,L-1 norm (GRMNMF) method, for exploring multi-view data. GRMNMF utilizes L-2,L-1 norm to calculate the error between the original data and the reconstructed data and simultaneously utilizes the geometrical structure of data space. A series of experiments were conducted in seven real data sets. The experimental results manifest the superiority of GRMNMF. |
Year | DOI | Venue |
|---|---|---|
2019 | 10.3233/FAIA190200 | Frontiers in Artificial Intelligence and Applications |
Keywords | DocType | Volume |
Nonnegative matrix factorization,Multi-view Clustering,L-2,L-1 norm,Graph Regularization | Conference | 320 |
ISSN | Citations | PageRank |
0922-6389 | 0 | 0.34 |
References | Authors | |
0 | 5 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Guowang Du | 1 | 1 | 1.36 |
| Lihua Zhou | 2 | 0 | 0.34 |
| Lizhen Wang | 3 | 0 | 0.34 |
| Qing Xiao | 4 | 7 | 3.52 |
| Hongmei Chen | 5 | 25 | 5.39 |