A Non-Negative Matrix Factorization Game

TL;DR

This paper introduces a game-theoretic approach to Non-Negative Matrix Factorization that improves scalability and parallelization, maintaining similar accuracy and convergence as traditional methods.

Contribution

It presents a novel game-theoretic formulation of NNMF that enhances scalability and parallelization without sacrificing performance.

Findings

Favorable scaling and parallelization properties

Reconstruction and convergence comparable to traditional algorithms

Applicable to various scientific and engineering data analysis tasks

Abstract

We present a novel game-theoretic formulation of Non-Negative Matrix Factorization (NNMF), a popular data-analysis method with many scientific and engineering applications. The game-theoretic formulation is shown to have favorable scaling and parallelization properties, while retaining reconstruction and convergence performance comparable to the traditional Multiplicative Updates algorithm.