Robustness Analysis of Hottopixx, a Linear Programming Model for Factoring Nonnegative Matrices

11/28/2012

∙

Although nonnegative matrix factorization (NMF) is NP-hard in general, it has been shown very recently that it is tractable under the assumption that the input nonnegative data matrix is close to being separable (separability requires that all columns of the input matrix belongs to the cone spanned by a small subset of these columns). Since then, several algorithms have been designed to handle this subclass of NMF problems. In particular, Bittorf, Recht, Ré and Tropp (`Factoring nonnegative matrices with linear programs', NIPS 2012) proposed a linear programming model, referred to as Hottopixx. In this paper, we provide a new and more general robustness analysis of their method. In particular, we design a provably more robust variant using a post-processing strategy which allows us to deal with duplicates and near duplicates in the dataset.

READ FULL TEXT

Robustness Analysis of Hottopixx, a Linear Programming Model for Factoring Nonnegative Matrices

Semidefinite Programming Based Preconditioning for More Robust Near-Separable Nonnegative Matrix Factorization

Robust Near-Separable Nonnegative Matrix Factorization Using Linear Optimization

Nonnegative partial s-goodness for the equivalence of a 0-1 linear program to weighted linear programming

Scalable methods for nonnegative matrix factorizations of near-separable tall-and-skinny matrices

On the Number of Many-to-Many Alignments of Multiple Sequences

First Order Methods for Robust Non-negative Matrix Factorization for Large Scale Noisy Data

Automatic Relevance Determination in Nonnegative Matrix Factorization with the β-Divergence

Robustness Analysis of Hottopixx, a Linear Programming Model for Factoring Nonnegative Matrices

Related Research

Semidefinite Programming Based Preconditioning for More Robust Near-Separable Nonnegative Matrix Factorization

Robust Near-Separable Nonnegative Matrix Factorization Using Linear Optimization

Nonnegative partial s-goodness for the equivalence of a 0-1 linear program to weighted linear programming

Scalable methods for nonnegative matrix factorizations of near-separable tall-and-skinny matrices

On the Number of Many-to-Many Alignments of Multiple Sequences

First Order Methods for Robust Non-negative Matrix Factorization for Large Scale Noisy Data

Automatic Relevance Determination in Nonnegative Matrix Factorization with the β-Divergence