Robust nonnegative matrix factorization using L21-norm

By Deguang Kong, Chris H. Q. Ding, Heng Huang. International Conference on Information and Knowledge Management, 2011

371 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Signal Processing

View Paper: https://www.semanticscholar.org/paper/8fee7b38358815e443f8316fa18768d76dba12e3

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Philip S. Yu
2. Charu C. Aggarwal
3. Tat-seng Chua
4. Xiangnan He
5. Peng Cui
6. Wenwu Zhu
7. Yi Yang
8. Hanwang Zhang

In Computer Vision and Pattern Recognition

1. Dacheng Tao
2. Shuicheng Yan
3. Qiang Yang
4. Yi Yang
5. Tat-seng Chua
6. Hanwang Zhang
7. Charu C. Aggarwal
8. Luc Van Gool

In Signal Processing

1. Charu C. Aggarwal
2. Philip S. Yu
3. Hyenkyun Woo
4. Ramakrishnan Kannan
5. Shuicheng Yan
6. Yi Yang
7. Qiang Yang
8. J. A. Carrasco-Ochoa

Citations in the corpus

• Doubly Aligned Incomplete Multi-view Clustering (2018, methodology): “…more than two incomplete views is proposed in [Shao et al. , 2015](MIC) by filling the missing instances with the average feature values in each incomplete view, then handling the problem with the help of weighted NMF and L 2 , 1 -Norm regularization [Kong et al. , 2011; Wu et al. , 2018].”
• Multiple Incomplete Views Clustering via Weighted Nonnegative Matrix Factorization with L2, 1 Regularization (2015, background): “Because of its robustness to noise and outliers, L2,1 has been widely used in many areas [11,13,18,21].”
• Deep Feature Selection: Theory and Application to Identify Enhancers and Promoters (2015, positive): “…j j w i w ij j in Equation (8) and Therefore, we cannot find a value of b to guarantee b k (cid:2) W Inspired by works of non-negative matrix factorization (NMF) (Nie et al., 2010; Kong et al., 2011), we learn that the row-wise sparseness of a matrix can be alternatively achieved using l 2,1 -norm.”
• Unsupervised Feature Selection Using Nonnegative Spectral Analysis (2012, methodology): “In that way, the proposed method is able to handle correlated and noisy features (Kong, Ding, and Huang 2011; Nie et al. 2010).”
• A review of unsupervised feature selection methods (2019, methodology): “Unlike the unsupervised feature selection methods above mentioned such as MCFS, UDFS, and NDFS, RUFS learns the pseudo cluster labels via local learning regularized robust nonnegative matrix factorization (Kong et al. 2011).”
• Outlier Detection for Text Data (2017, methodology): “Also, Kong et al [14] propose a robust formulation of NMF using (cid:96) 2 , 1 -norm loss function for data with noises.”
• Social contextual recommendation (2012, background): “The matrix approximations all focus on representing the user-item rating matrix with low-dimensional latent vectors [24, 15, 12].”
• Robust (Semi) Nonnegative Graph Embedding (2014, background): “Particularly, ‖ · ‖F is the Frobenius norm and ‖ · ‖1 is the 1-norm which is the sum of the absolute value of all the entries in the matrix (or vector).”