Robust Nonnegative Matrix Factorization with Ordered Structure Constraints

Authors: Wang, J., Tian, F., Liu, C.H., Yu, H.C., Wang, X. and Tang, X.C.

http://eprints.bournemouth.ac.uk/29417/

Start date: 14 May 2017

Nonnegative matrix factorization (NMF) as a popular technique to find parts- based representations of nonnegative data has been widely used in real-world applications. Often the data which these applications process, such as motion sequences and video clips, are with ordered structure, i.e., consecutive neighbouring data samples are very likely share similar features unless a sudden change occurs. Therefore, traditional NMF assumes the data samples and features to be independently distributed, making it not proper for the analysis of such data. In this paper, we propose an ordered robust NMF (ORNMF) by capturing the embedded ordered structure to improve the accuracy of data representation. With a novel neighbour penalty term, ORNMF enforces the similarity of neighbouring data. ORNMF also adopts the $L_{2,1}$-norm based loss function to improve its robustness against noises and outliers. A new iterative updating optimization algorithm is derived to solve ORNMF's objective function. The proofs of the convergence and correctness of the scheme are also presented. Experiments on both synthetic and real-world datasets have demonstrated the effectiveness of ORNMF.

This data was imported from Scopus:

Authors: Wang, J., Tian, F., Liu, C.H., Yu, H., Wang, X. and Tang, X.

http://eprints.bournemouth.ac.uk/29417/

Journal: Proceedings of the International Joint Conference on Neural Networks

Volume: 2017-May

Pages: 478-485

ISBN: 9781509061815

DOI: 10.1109/IJCNN.2017.7965892

© 2017 IEEE. Nonnegative matrix factorization (NMF) as a popular technique to find parts-based representations of nonnegative data has been widely used in real-world applications. Often the data which these applications process, such as motion sequences and video clips, are with ordered structure, i.e., consecutive neighbouring data samples are very likely share similar features unless a sudden change occurs. Therefore, traditional NMF assumes the data samples and features to be independently distributed, making it not proper for the analysis of such data. In this paper, we propose an ordered robust NMF (ORNMF) by capturing the embedded ordered structure to improve the accuracy of data representation. With a novel neighbour penalty term, ORNMF enforces the similarity of neighbouring data. ORNMF also adopts the L2,1-norm based loss function to improve its robustness against noises and outliers. A new iterative updating optimization algorithm is derived to solve ORNMF's objective function. The proofs of the convergence and correctness of the scheme are also presented. Experiments on both synthetic and real-world datasets have demonstrated the effectiveness of ORNMF.

This data was imported from Web of Science (Lite):

Authors: Wang, J., Tian, F., Liu, C.H., Yu, H., Wang, X., Tang, X. and IEEE

http://eprints.bournemouth.ac.uk/29417/

Journal: 2017 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN)

Pages: 478-485

ISSN: 2161-4393

The data on this page was last updated at 04:51 on November 17, 2018.