A novel framework for the NMF methods with experiments to unmixing signals and feature representation

Yueyang Teng, Yu-Dong Yao, Shouliang Qi, Chen Li, Lisheng Xu, Wei Qian, Fenglei Fan, Ge Wang

Research output: Contribution to journalArticle

Abstract

Non-negative matrix factorization (NMF) can be used in clustering, feature representation or blind source separation. Many NMF methods have been developed including least squares (LS) error, Kullback–Leibler (KL) divergence, Itakura–Saito (IS) divergence, Bregman-divergence, α-divergence, β-divergence, γ-divergence, convex, constrained, graph-regularized NMFs. The main contribution of this paper is to develop a framework to generalize the existing NMF methods and also provide new NMF methods. This paper constructs a general optimization model and develops a generic updating rule with a simple structure using a surrogate function, which possesses similar properties as the standard NMF methods. The experimental results, obtained using several standard databases, demonstrate the power of the work in which some new methods provide performance superior to that of the other existing methods.

Original languageEnglish (US)
Pages (from-to)205-218
Number of pages14
JournalJournal of Computational and Applied Mathematics
Volume362
DOIs
StatePublished - Dec 15 2019

Fingerprint

Factorization Method
Matrix Factorization
Matrix Method
Factorization
Divergence
Experiment
Experiments
Bregman Divergence
Non-negative Matrix Factorization
Kullback-Leibler Divergence
Blind Source Separation
Optimization Model
Blind source separation
Updating
Least Squares
Clustering
Generalise
Framework
Experimental Results
Graph in graph theory

All Science Journal Classification (ASJC) codes

  • Computational Mathematics
  • Applied Mathematics

Cite this

Teng, Yueyang ; Yao, Yu-Dong ; Qi, Shouliang ; Li, Chen ; Xu, Lisheng ; Qian, Wei ; Fan, Fenglei ; Wang, Ge. / A novel framework for the NMF methods with experiments to unmixing signals and feature representation. In: Journal of Computational and Applied Mathematics. 2019 ; Vol. 362. pp. 205-218.
@article{952eaa776a7b4f7f9a03b9c4cbd5e4d2,
title = "A novel framework for the NMF methods with experiments to unmixing signals and feature representation",
abstract = "Non-negative matrix factorization (NMF) can be used in clustering, feature representation or blind source separation. Many NMF methods have been developed including least squares (LS) error, Kullback–Leibler (KL) divergence, Itakura–Saito (IS) divergence, Bregman-divergence, α-divergence, β-divergence, γ-divergence, convex, constrained, graph-regularized NMFs. The main contribution of this paper is to develop a framework to generalize the existing NMF methods and also provide new NMF methods. This paper constructs a general optimization model and develops a generic updating rule with a simple structure using a surrogate function, which possesses similar properties as the standard NMF methods. The experimental results, obtained using several standard databases, demonstrate the power of the work in which some new methods provide performance superior to that of the other existing methods.",
author = "Yueyang Teng and Yu-Dong Yao and Shouliang Qi and Chen Li and Lisheng Xu and Wei Qian and Fenglei Fan and Ge Wang",
year = "2019",
month = "12",
day = "15",
doi = "https://doi.org/10.1016/j.cam.2019.05.010",
language = "English (US)",
volume = "362",
pages = "205--218",
journal = "Journal of Computational and Applied Mathematics",
issn = "0377-0427",
publisher = "Elsevier",

}

A novel framework for the NMF methods with experiments to unmixing signals and feature representation. / Teng, Yueyang; Yao, Yu-Dong; Qi, Shouliang; Li, Chen; Xu, Lisheng; Qian, Wei; Fan, Fenglei; Wang, Ge.

In: Journal of Computational and Applied Mathematics, Vol. 362, 15.12.2019, p. 205-218.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A novel framework for the NMF methods with experiments to unmixing signals and feature representation

AU - Teng, Yueyang

AU - Yao, Yu-Dong

AU - Qi, Shouliang

AU - Li, Chen

AU - Xu, Lisheng

AU - Qian, Wei

AU - Fan, Fenglei

AU - Wang, Ge

PY - 2019/12/15

Y1 - 2019/12/15

N2 - Non-negative matrix factorization (NMF) can be used in clustering, feature representation or blind source separation. Many NMF methods have been developed including least squares (LS) error, Kullback–Leibler (KL) divergence, Itakura–Saito (IS) divergence, Bregman-divergence, α-divergence, β-divergence, γ-divergence, convex, constrained, graph-regularized NMFs. The main contribution of this paper is to develop a framework to generalize the existing NMF methods and also provide new NMF methods. This paper constructs a general optimization model and develops a generic updating rule with a simple structure using a surrogate function, which possesses similar properties as the standard NMF methods. The experimental results, obtained using several standard databases, demonstrate the power of the work in which some new methods provide performance superior to that of the other existing methods.

AB - Non-negative matrix factorization (NMF) can be used in clustering, feature representation or blind source separation. Many NMF methods have been developed including least squares (LS) error, Kullback–Leibler (KL) divergence, Itakura–Saito (IS) divergence, Bregman-divergence, α-divergence, β-divergence, γ-divergence, convex, constrained, graph-regularized NMFs. The main contribution of this paper is to develop a framework to generalize the existing NMF methods and also provide new NMF methods. This paper constructs a general optimization model and develops a generic updating rule with a simple structure using a surrogate function, which possesses similar properties as the standard NMF methods. The experimental results, obtained using several standard databases, demonstrate the power of the work in which some new methods provide performance superior to that of the other existing methods.

UR - http://www.scopus.com/inward/record.url?scp=85066744499&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85066744499&partnerID=8YFLogxK

U2 - https://doi.org/10.1016/j.cam.2019.05.010

DO - https://doi.org/10.1016/j.cam.2019.05.010

M3 - Article

VL - 362

SP - 205

EP - 218

JO - Journal of Computational and Applied Mathematics

JF - Journal of Computational and Applied Mathematics

SN - 0377-0427

ER -