A variational approach to stable principal component pursuit

TL;DR

This paper presents a new convex variational method for stable principal component pursuit that improves scalability and parameter tuning, demonstrated through synthetic and real data experiments.

Contribution

It introduces a novel convex formulation and a variational framework for SPCP, accelerated with quasi-Newton methods, enhancing practical applicability.

Findings

Outperforms classical SPCP in scalability

Offers easier parameter selection

Effective on synthetic and real datasets

Abstract

We introduce a new convex formulation for stable principal component pursuit (SPCP) to decompose noisy signals into low-rank and sparse representations. For numerical solutions of our SPCP formulation, we first develop a convex variational framework and then accelerate it with quasi-Newton methods. We show, via synthetic and real data experiments, that our approach offers advantages over the classical SPCP formulations in scalability and practical parameter selection.