Bidirectional Nonnegative Deep Model and Its Optimization in Learning
Joint Authors
Yu, Hong
Zeng, Xianhua
He, Zhengyi
Qu, Shengwei
Source
Issue
Vol. 2016, Issue 2016 (31 Dec. 2016), pp.1-8, 8 p.
Publisher
Hindawi Publishing Corporation
Publication Date
2016-11-16
Country of Publication
Egypt
No. of Pages
8
Main Subjects
Abstract EN
Nonnegative matrix factorization (NMF) has been successfully applied in signal processing as a simple two-layer nonnegative neural network.
Projective NMF (PNMF) with fewer parameters was proposed, which projects a high-dimensional nonnegative data onto a lower-dimensional nonnegative subspace.
Although PNMF overcomes the problem of out-of-sample of NMF, it does not consider the nonlinear characteristic of data and is only a kind of narrow signal decomposition method.
In this paper, we combine the PNMF with deep learning and nonlinear fitting to propose a bidirectional nonnegative deep learning (BNDL) model and its optimization learning algorithm, which can obtain nonlinear multilayer deep nonnegative feature representation.
Experiments show that the proposed model can not only solve the problem of out-of-sample of NMF but also learn hierarchical nonnegative feature representations with better clustering performance than classical NMF, PNMF, and Deep Semi-NMF algorithms.
American Psychological Association (APA)
Zeng, Xianhua& He, Zhengyi& Yu, Hong& Qu, Shengwei. 2016. Bidirectional Nonnegative Deep Model and Its Optimization in Learning. Journal of Optimization،Vol. 2016, no. 2016, pp.1-8.
https://search.emarefa.net/detail/BIM-1110151
Modern Language Association (MLA)
Zeng, Xianhua…[et al.]. Bidirectional Nonnegative Deep Model and Its Optimization in Learning. Journal of Optimization No. 2016 (2016), pp.1-8.
https://search.emarefa.net/detail/BIM-1110151
American Medical Association (AMA)
Zeng, Xianhua& He, Zhengyi& Yu, Hong& Qu, Shengwei. Bidirectional Nonnegative Deep Model and Its Optimization in Learning. Journal of Optimization. 2016. Vol. 2016, no. 2016, pp.1-8.
https://search.emarefa.net/detail/BIM-1110151
Data Type
Journal Articles
Language
English
Notes
Includes bibliographical references
Record ID
BIM-1110151