Approximate Nonnegative Matrix Factorization via Alternating Minimization

Publication date: 13 February 2004

Abstract: In this paper we consider the Nonnegative Matrix Factorization (NMF) problem: given an (elementwise) nonnegative matrix

V i n R_{+}^{m i m e s n}

find, for assigned

k

, nonnegative matrices

W i n R_{+}^{m i m e s k}

and

H i n R_{+}^{k i m e s n}

such that

V = W H

. 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

k

, the factorization

W H

closest to

V

in I-divergence. An iterative algorithm, EM like, for the construction of the best pair

(W, H)

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.

Mathematics Subject Classification ID

Stochastic systems in control theory (general) (93E03)

This page was built for publication: Approximate Nonnegative Matrix Factorization via Alternating Minimization