A Unified Framework for Sparse Non-Negative Least Squares using Multiplicative Updates and the Non-Negative Matrix Factorization Problem

TL;DR

This paper introduces a unified, efficient framework for solving sparse non-negative least squares problems and extends it to sparse non-negative matrix factorization, with theoretical guarantees and practical performance validation.

Contribution

The authors propose a novel unified framework for S-NNLS and S-NMF using rectified power exponential priors, including multiplicative update algorithms with convergence guarantees.

Findings

Algorithms converge to stationary points of the objective function.

Framework encompasses many existing S-NNLS and S-NMF algorithms.

Validated performance on synthetic and real-world datasets.

Abstract

We study the sparse non-negative least squares (S-NNLS) problem. S-NNLS occurs naturally in a wide variety of applications where an unknown, non-negative quantity must be recovered from linear measurements. We present a unified framework for S-NNLS based on a rectified power exponential scale mixture prior on the sparse codes. We show that the proposed framework encompasses a large class of S-NNLS algorithms and provide a computationally efficient inference procedure based on multiplicative update rules. Such update rules are convenient for solving large sets of S-NNLS problems simultaneously, which is required in contexts like sparse non-negative matrix factorization (S-NMF). We provide theoretical justification for the proposed approach by showing that the local minima of the objective function being optimized are sparse and the S-NNLS algorithms presented are guaranteed to converge to a set of stationary points of the objective function. We then extend our framework to S-NMF, showing that our framework leads to many well known S-NMF algorithms under specific choices of prior and providing a guarantee that a popular subclass of the proposed algorithms converges to a set of stationary points of the objective function. Finally, we study the performance of the proposed approaches on synthetic and real-world data.