Fast Local Learning Regularized Nonnegative Matrix Factorization

被引：0

作者：

Jiang, Jiaojiao ^{[1
]}

Zhang, Haibin ^{[1
]}

Xue, Yi ^{[1
]}

机构：

[1] Beijing Univ Technol, Coll Appl Sci, Beijing 100124, Peoples R China

来源：

ADVANCES IN COMPUTATIONAL ENVIRONMENT SCIENCE | 2012年 / 142卷

关键词：

local learning regularization; Nonnegative matrix factorization; NMF convergent speed;

D O I：

暂无

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

Non-negative matrix factorization (NMF) is a recently developed technique for finding parts-based, linear representations of non-negative data. In this paper, we present a fast algorithm to solve local learning regularized nonnegative matrix factorization. We consider not only the local learning, but also its convergence speed. Experiments on many benchmark data sets demonstrate that the proposed method outperforms the local learning regularized NMF in convergence speed.

引用

页码：67 / 75

页数：9

共 10 条

[1] LOCAL LEARNING ALGORITHMS [J].

BOTTOU, L ;

VAPNIK, V .

NEURAL COMPUTATION, 1992, 4 (06) :888-900

[2]

DING C., 2008, IEEE Transactions on Pattern Analysis and Machine Intelligence, V99

[3]

Ding C.H., 2006, PROC 12 ACM SIGKDD I, P126

[4]

Gu QQ, 2009, 21ST INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE (IJCAI-09), PROCEEDINGS, P1046

[5]

Hastie T., 2001, ELEMENTS STAT LEARNI

[6]

Hoyer PO, 2004, J MACH LEARN RES, V5, P1457

[7]

Lee DD, 2001, ADV NEUR IN, V13, P556

[8] FastNMF: highly efficient monotonic fixed-point nonnegative matrix factorization algorithm with good applicability [J].

Li, Le ;

Zhang, Yu-Jin .

JOURNAL OF ELECTRONIC IMAGING, 2009, 18 (03)

[9]

Li SZ, 2001, PROC CVPR IEEE, P207

[10]

Scholkopf B, 2002, Encyclopedia of Biostatistics

← 1 →