A nonnegative projection based algorithm for low-rank nonnegative matrix approximation

Peitao Wang, Zhaoshui He, Kan Xie, Junbin Gao, Michael Antolovich

Research output: Book chapter/Published conference paperConference paperpeer-review

1 Citation (Scopus)


Nonnegative matrix factorization/approximation (NMF/NMA) is a widely used method for data analysis. So far, many multiplicative update algorithms have been developed for NMF. In this paper, we propose a nonnegative projection based NMF algorithm, which is different from the conventional multiplicative update NMF algorithms and decreases the objective function by performing Procrustes rotation and nonnegative projection alternately. The experiment results demonstrate that the new algorithm converges much faster than traditional ones.
Original languageEnglish
Title of host publicationNeural information processing
Subtitle of host publication24th International Conference, ICONIP 2017, Guangzhou, China, November 14-18, 2017, Proceedings, Part I
EditorsDerong Liu, Shengli Xie, Yuanqing Li, Dongbin Zhao, El-Sayed M. El-Alfy
PublisherSpringer-Verlag London Ltd.
Number of pages8
ISBN (Electronic)9783319700878
ISBN (Print)9783319700861
Publication statusPublished - 24 Oct 2017
Event24th International Conference on Neural Information Processing: ICONIP 2017 - Ramada Pearl Hotel, Guangzhou, China
Duration: 14 Nov 201718 Nov 2017
http://www.conference123.org/iconip2017/index.html (Conference website)
https://link.springer.com/book/10.1007/978-3-319-70087-8#about (Conference proceedings)

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10634 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference24th International Conference on Neural Information Processing
OtherThe 24th International Conference on Neural Information Processing will be held in Guangzhou, China, November 14–18, 2017. It is an annual event, organized since 1994 by the Asia Pacific Neural Network Society (APNNS, previously APNNA).
Internet address


Dive into the research topics of 'A nonnegative projection based algorithm for low-rank nonnegative matrix approximation'. Together they form a unique fingerprint.

Cite this