A proximal point algorithm for sequential feature extraction applications

TL;DR

This paper introduces a proximal point algorithm for the LAROS problem, enabling sequential feature extraction from data, with a new stopping criterion backed by theoretical guarantees, demonstrated on image datasets.

Contribution

The paper presents a novel proximal point algorithm for LAROS with a duality-based stopping criterion and applies it to feature extraction in images.

Findings

Effective extraction of common features from images

The algorithm converges with the proposed stopping criterion

Demonstrated on two image databases

Abstract

We propose a proximal point algorithm to solve LAROS problem, that is the problem of finding a "large approximately rank-one submatrix". This LAROS problem is used to sequentially extract features in data. We also develop a new stopping criterion for the proximal point algorithm, which is based on the duality conditions of \eps-optimal solutions of the LAROS problem, with a theoretical guarantee. We test our algorithm with two image databases and show that we can use the LAROS problem to extract appropriate common features from these images.