On trivial solution and scale transfer problems in graph regularized NMF

Quanquan Gu, Chris Ding, Jiawei Han

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Combining graph regularization with nonnegative matrix (tri-)factorization (NMF) has shown great performance improvement compared with traditional nonnegative matrix (tri-)factorization models due to its ability to utilize the geometric structure of the documents and words. In this paper, we show that these models are not well-defined and suffering from trivial solution and scale transfer problems. In order to solve these common problems, we propose two models for graph regularized nonnegative matrix (tri-)factorization, which can be applied for document clustering and co-clustering respectively. In the proposed models, a Normalized Cut-like constraint is imposed on the cluster assignment matrix to make the optimization problem well-defined. We derive a multiplicative updating algorithm for the proposed models, and prove its convergence. Experiments of clustering and coclustering on benchmark text data sets demonstrate that the proposed models outperform the original models as well as many other state-of-the-art clustering methods.

Original languageEnglish (US)
Title of host publicationIJCAI 2011 - 22nd International Joint Conference on Artificial Intelligence
Pages1288-1293
Number of pages6
DOIs
StatePublished - Dec 1 2011
Event22nd International Joint Conference on Artificial Intelligence, IJCAI 2011 - Barcelona, Catalonia, Spain
Duration: Jul 16 2011Jul 22 2011

Publication series

NameIJCAI International Joint Conference on Artificial Intelligence
ISSN (Print)1045-0823

Other

Other22nd International Joint Conference on Artificial Intelligence, IJCAI 2011
Country/TerritorySpain
CityBarcelona, Catalonia
Period7/16/117/22/11

ASJC Scopus subject areas

  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'On trivial solution and scale transfer problems in graph regularized NMF'. Together they form a unique fingerprint.

Cite this