第4 期 电 子 学 报 Vol. 36 No. 4
2008年4 月 ACTA ELECTRONICA SINICA Apr. 2008
非负矩阵分解算法综述
1, 2 1, 2
,
( 11 清华信息科学与技术国家实验室, 北京 100084; 21 清华大学电子工程 , 北京 100084)
: 本文介绍了非负矩阵分解(Non2negative Matrix Factorization,NMF) 的基本原理和性质, 将现有NMF 算法分
为了基于基本NMF 模型的算法和基于改进NMF 模型的算法两大类, 在此基础上较为 统地分析、总结和比较了它们
的构造原则、应用特点以及存在的问题, 最后预测和分析了未来NMF 算法研究的可能方向.
: 非负矩阵分解; 多元数据描述; 特征提取
: TP391 : A : 03 222112 (2008) 042 0 3 2 0
A Survey on Algorithms of Non2Negative Matrix Factorization
1, 2 1, 2
LI Le , ZHANG Yu2jin
(11Tsinghua National Laboratory for Information Scienceand Technology, Tsinghua University, Beijing 100084, China;
21 epartment of ElectronicEngineering, Tsinghua University,Beijing 100084, China)
Ab stract: The fundamentals and properties of non2negative matrix factorization (NMF) are introduced, and available NMF
algorithms are classified into two categories:basic NMF mode2l based algorithms and improved NMF model2based algorithms. Based
on these,the design principles, application characteristics, and existing problems of the algorithms are systematically discussed. Be2
sides, some open problems in the development of NMF algorithms are presented and analyzed.
Ke y word s : non2negative matrix factorization;multivariate data representation;feature extraction
[ ]
种有效数据描述形式 .