Nonnegative low rank matrix approximation for nonnegative matrices
From MaRDI portal
Publication:1985392
DOI10.1016/J.AML.2020.106300zbMATH Open1436.65056arXiv1912.06836OpenAlexW3008742172MaRDI QIDQ1985392FDOQ1985392
Authors: G. J. Song, Michael Ng 
Publication date: 7 April 2020
Published in: Applied Mathematics Letters (Search for Journal in Brave)
Abstract: This paper describes a new algorithm for computing Nonnegative Low Rank Matrix (NLRM) approximation for nonnegative matrices. Our approach is completely different from classical nonnegative matrix factorization (NMF) which has been studied for more than twenty five years. For a given nonnegative matrix, the usual NMF approach is to determine two nonnegative low rank matrices such that the distance between their product and the given nonnegative matrix is as small as possible. However, the proposed NLRM approach is to determine a nonnegative low rank matrix such that the distance between such matrix and the given nonnegative matrix is as small as possible. There are two advantages. (i) The minimized distance by the proposed NLRM method can be smaller than that by the NMF method, and it implies that the proposed NLRM method can obtain a better low rank matrix approximation. (ii) Our low rank matrix admits a matrix singular value decomposition automatically which provides a significant index based on singular values that can be used to identify important singular basis vectors, while this information cannot be obtained in the classical NMF. The proposed NLRM approximation algorithm was derived using the alternating projection on the low rank matrix manifold and the non-negativity property. Experimental results are presented to demonstrate the above mentioned advantages of the proposed NLRM method compared the NMF method.
Full work available at URL: https://arxiv.org/abs/1912.06836
Recommendations
- An approximate augmented Lagrangian method for nonnegative low-rank matrix approximation
- Structured low rank approximation of non-negative matrices
- Low-Dimensional Polytope Approximation and Its Applications to Nonnegative Matrix Factorization
- scientific article; zbMATH DE number 7404610
- Nonnegative Matrix Factorization Based on Alternating Nonnegativity Constrained Least Squares and Active Set Method
Eigenvalues, singular values, and eigenvectors (15A18) Numerical computation of eigenvalues and eigenvectors of matrices (65F15) Positive matrices and their generalizations; cones of matrices (15B48) Numerical methods for low-rank matrix approximation; matrix compression (65F55)
Cites Work
- Title not available (Why is that?)
- Title not available (Why is that?)
- Projected Gradient Methods for Nonnegative Matrix Factorization
- Learning the parts of objects by non-negative matrix factorization
- Introduction to Smooth Manifolds
- Low-rank matrix completion by Riemannian optimization
- Hierarchical ALS Algorithms for Nonnegative Matrix and 3D Tensor Factorization
- Introducing a weighted non-negative matrix factorization for image classification
- Alternating projections on nontangential manifolds
- Robust and Non-Negative Collective Matrix Factorization for Text-to-Image Transfer Learning
- SNMFCA: Supervised NMF-Based Image Classification and Annotation
Cited In (19)
- A random sampling algorithm for fully-connected tensor network decomposition with applications
- Title not available (Why is that?)
- An approximate augmented Lagrangian method for nonnegative low-rank matrix approximation
- Quasi-orthogonalization for alternating non-negative tensor factorization
- On reduced rank nonnegative matrix factorization for symmetric nonnegative matrices
- Sketching for a low-rank nonnegative matrix approximation: numerical study
- Non-negative low-rank approximations for multi-dimensional arrays on statistical manifold
- Nonnegative low rank tensor approximations with multidimensional image applications
- Low rank approximation of the symmetric positive semidefinite matrix
- A remark on low rank matrix recovery and noncommutative Bernstein type inequalities
- Sequential Quadratic Optimization for Nonlinear Optimization Problems on Riemannian Manifolds
- Randomized low rank approximation for nonnegative pure quaternion matrices
- Low-rank nonnegative tensor approximation via alternating projections and sketching
- Nonnegative tensor train for the multicomponent Smoluchowski equation
- Riemannian Interior Point Methods for Constrained Optimization on Manifolds
- Low-Dimensional Polytope Approximation and Its Applications to Nonnegative Matrix Factorization
- Study of performance of low-rank nonnegative tensor factorization methods
- A Nonlinear Matrix Decomposition for Mining the Zeros of Sparse Data
- Computing Closest Stable Nonnegative Matrix
Uses Software
This page was built for publication: Nonnegative low rank matrix approximation for nonnegative matrices
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1985392)
