A symmetric rank-one quasi-Newton method for nonnegative matrix factorization

DOI10.1155/2014/846483zbMATH Open1298.65101arXiv1305.5829OpenAlexW1963795260WikidataQ59048704 ScholiaQ59048704MaRDI QIDQ469969FDOQ469969

Authors: Shu-Zhen Lai, Hou-Biao Li, Zu-Tao Zhang 

Publication date: 11 November 2014

Published in: ISRN Applied Mathematics (Search for Journal in Brave)

Abstract: As we all known, the nonnegative matrix factorization (NMF) is a dimension reduction method that has been widely used in image processing, text compressing and signal processing etc. In this paper, an algorithm for nonnegative matrix approximation is proposed. This method mainly bases on the active set and the quasi-Newton type algorithm, by using the symmetric rank-one and negative curvature direction technologies to approximate the Hessian matrix. Our method improves the recent results of those methods in [Pattern Recognition, 45(2012)3557-3565; SIAM J. Sci. Comput., 33(6)(2011)3261-3281; Neural Computation, 19(10)(2007)2756-2779, etc.]. Moreover, the object function decreases faster than many other NMF methods. In addition, some numerical experiments are presented in the synthetic data, imaging processing and text clustering. By comparing with the other six nonnegative matrix approximation methods, our experiments confirm to our analysis.

Full work available at URL: https://arxiv.org/abs/1305.5829

Recommendations

• Image processing using Newton-based algorithm of nonnegative matrix factorization
• Fast nonnegative matrix factorization: an active-set-like method and comparisons
• Adaptive computation of the symmetric nonnegative matrix factorization (SymNMF)
• Efficient nonnegative matrix factorization via projected Newton method
• scientific article; zbMATH DE number 7404610

Mathematics Subject Classification ID

Numerical mathematical programming methods (65K05) Nonlinear programming (90C30)

Cites Work

• Title not available (Why is that?)
• Algorithms and applications for approximate nonnegative matrix factorization
• Projected Gradient Methods for Nonnegative Matrix Factorization
• Learning the parts of objects by non-negative matrix factorization
• Title not available (Why is that?)
• Fast Nonnegative Matrix Factorization: An Active-Set-Like Method and Comparisons
• Projected Newton Methods for Optimization Problems with Simple Constraints
• Nonnegative Matrix Factorization Based on Alternating Nonnegativity Constrained Least Squares and Active Set Method
• Fast Projection‐Based Methods for the Least Squares Nonnegative Matrix Approximation Problem
• Efficient nonnegative matrix factorization via projected Newton method
• A symmetric rank-one quasi-Newton line-search method using negative curvature directions

Cited In (6)

• Title not available (Why is that?)
• On reduced rank nonnegative matrix factorization for symmetric nonnegative matrices
• Non-negative low-rank approximations for multi-dimensional arrays on statistical manifold
• Iterative algorithm for the Q-weighted nonnegative matrix factorization
• Computing symmetric nonnegative rank factorizations
• Active set type algorithms for nonnegative matrix factorization in hyperspectral unmixing

This page was built for publication: A symmetric rank-one quasi-Newton method for nonnegative matrix factorization

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q469969)