Iterative algorithm for the Q-weighted nonnegative matrix factorization. Issue 11 (18th August 2021)

• Record Type: Journal Article
• Authors: Li, Chun-Mei; Lu, Lin-Zhang; Chen, Zhen
• Is Part Of: Linear & multilinear algebra. Volume 69:Issue 11(2021)
• Page Start: 2130
• Page End: 2142
• Publication Date: 2021-08-18
• Subjects: Weighted nonnegative matrix factorization -- Q-weighted norm -- Iterative algorithm; 15A23 -- 41A36 -- 52B11; Algebras, Linear -- Periodicals; Multilinear algebra -- Periodicals; 512.505
• DOI: 10.1080/03081087.2019.1663138
• Languages: English
• ISSNs: 0308-1087

Abstract

Abstract : Nonnegative matrix factorization, which decomposes a target matrix into the product of two matrices with nonnegative elements, has been widely used in various fields of science, engineering and technology. In this paper, we consider the more general Q-weighted nonnegative matrix factorization (QWNMF) problem. By using the additive representation of the Q-weighted norm, the QWNMF problem is transformed into an unconstraint optimization problem, and then a new iterative algorithm is designed to solve it. The numerical analysis of this algorithm is also given. Numerical examples show that the new method is feasible and effective.