[1]YANG Zhi-jun,YE Dong-yi.A dynamic learning algorithm based on nonnegative matrix factorization[J].CAAI Transactions on Intelligent Systems,2010,5(4):320-326.
Copy

A dynamic learning algorithm based on nonnegative matrix factorization

CAAI Transactions on Intelligent Systems[ISSN 1673-4785/CN 23-1538/TP] Volume: 5 Number of periods: 2010 4 Page number: 320-326 Column: 学术论文—人工智能基础 Public date: 2010-08-25
Title:
A dynamic learning algorithm based on nonnegative matrix factorization
Author(s):
YANG Zhi-jun; YE Dong-yi
College of Mathematics and Computer Science, Fuzhou University, Fuzhou 350108, China
Keywords:
nonnegative matrix factorization;  dynamic learning;  initialization;  error criteria
CLC:
TP181
DOI:
-
Abstract:
To improve the performance of the incremental nonnegative matrix factorization algorithm, error estimation criteria for judging the effectiveness of the incremental algorithm was presented. Then, a new dynamic nonnegative matrix factorization algorithm was proposed whereby incremental factorization was initialized with the already factorized matrices before adding new samples. Experimental results on a number of data sets showed that the proposed algorithm is capable of instantly updating both the base matrix and the code matrix. Another benefit of the method is that the computational complexity is relatively low. The proposed algorithm can also identify noise points when dealing with dynamic data. So it is a feasible and effective dynamic factorization algorithm.
References:
［1］LEE D D, SEUNG H S. Learning the parts of objects by nonnegative matrix factorization［J］. Nature, 1999(401): 788-791.
［2］XU Wei, LU Xin, GONG Yihong. Document clustering based on nonnegative matrix factorization［C］//Proceedings of the 26th Annual International ACM SIGIR Conference, Toronto, Canada, 2003: 267-273.
［3］GUILLAMET D, VITRIA J. Nonnegative matrix factorization for face recognition［C］//Proc Conf Topics in Artificial Intelligence. Alberta, Canada, 2002: 336-344. 
［4］VIRTANEN T. Monaural sound source separation by nonnegative matrix factorization with temporal continuity and sparseness criteria［J］. IEEE Transactions on Audio, Speech, and Language Processing, 2007, 15(3): 1066-1074. 
［5］BUCAK S S, GUNSEL B, GURSOY O. Incremental nonnegative matrix factorization for dynamic background modeling［C］//ICEIS International Workshop on Pattern Recognition in Information Systems. Funchal, Portugal, 2007: 107-116.
［6］CAO Bin, SHEN Dou, SUN Jiantao, et al. Detect and track latent factors with online nonnegative matrix factorization［C］//The 20th International Joint Conference on Artificial Intelligence. Hyderabad, India, 2007: 2689-2694. 
［7］CHEN W S, PAN B B, FANG B, et al. A novel constraint nonnegative matrix factorization criterion based incremental learning in face recognition［C］//Proceedings of International Conference on Wavelet Analysis and Pattern Recognition. Hong Kong, China, 2008: 292-297. 
［8］ PETS Video Database［EB/OL］. http://ftp.pets.rdg.ac.uk/.
［9］DONOHO D， Stodden V. When does nonnegative matrix factorization give a correct decomposition into parts?［C］//Proceedings of the Seventeenth Annual Conference on Neural Information Processing Systems. Vancouver and Whistler, Canada, 2003: 101-108. 
［10］LI T, DING C. The relationships among various nonnegative matrix factorization methods for clustering［C］//Sixth IEEE International Conference on Data Mining. Hong Kong, China, 2006: 362-371.
Similar References:

Memo

-

Last Update: 2010-09-20

Copyright © CAAI Transactions on Intelligent Systems