Image classification with nonnegative matrix factorization based on spectral projected gradient

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

6 Citations (Scopus)

Abstract

Nonnegative Matrix Factorization (NMF) is a key tool for model dimensionality reduction in supervised classification. Several NMF algorithms have been developed for this purpose. In a majority of them, the training process is improved by using discriminant or nearest-neighbor graph-based constraints that are obtained from the knowledge on class labels of training samples. The constraints are usually incorporated to NMF algorithms by l2-weighted penalty terms that involve formulating a large-size weighting matrix. Using the Newton method for updating the latent factors, the optimization problems in NMF become large-scale. However, the computational problem can be considerably alleviated if the modified Spectral Projected Gradient (SPG) that belongs to a class of quasi-Newton methods is used. The simulation results presented for the selected classification problems demonstrate the high efficiency of the proposed method.

Original languageEnglish
Title of host publicationArtificial Neural Networks - Methods and Applications in Bio-/Neuroinformatics
PublisherSpringer Verlag
Pages31-50
Number of pages20
ISBN (Print)9783319099026
DOIs
Publication statusPublished - 2015
Externally publishedYes
Event23rd International Conference on Artificial Neural Networks, ICANN 2013 - Sofia, Bulgaria
Duration: 10 Sep 201313 Sep 2013

Publication series

NameArtificial Neural Networks - Methods and Applications in Bio-/Neuroinformatics

Conference

Conference23rd International Conference on Artificial Neural Networks, ICANN 2013
Country/TerritoryBulgaria
CitySofia
Period10/09/1313/09/13

Fingerprint

Dive into the research topics of 'Image classification with nonnegative matrix factorization based on spectral projected gradient'. Together they form a unique fingerprint.

Cite this