Approximate Nonnegative Matrix Factorization via Alternating Minimization
Abstract
In this paper we consider the Nonnegative Matrix Factorization (NMF) problem: given an (elementwise) nonnegative matrix find, for assigned , nonnegative matrices and such that . Exact, non trivial, nonnegative factorizations do not always exist, hence it is interesting to pose the approximate NMF problem. The criterion which is commonly employed is I-divergence between nonnegative matrices. The problem becomes that of finding, for assigned , the factorization closest to in I-divergence. An iterative algorithm, EM like, for the construction of the best pair has been proposed in the literature. In this paper we interpret the algorithm as an alternating minimization procedure \`a la Csisz\'ar-Tusn\'ady and investigate some of its stability properties. NMF is widespreading as a data analysis method in applications for which the positivity constraint is relevant. There are other data analysis methods which impose some form of nonnegativity: we discuss here the connections between NMF and Archetypal Analysis. An interesting system theoretic application of NMF is to the problem of approximate realization of Hidden Markov Models.
Keywords
Cite
@article{arxiv.math/0402229,
title = {Approximate Nonnegative Matrix Factorization via Alternating Minimization},
author = {Lorenzo Finesso and Peter Spreij},
journal= {arXiv preprint arXiv:math/0402229},
year = {2007}
}