Analysis of the Self Projected Matching Pursuit Algorithm

TL;DR

This paper analyzes the convergence and numerical properties of the Self Projected Matching Pursuit algorithm, a memory-efficient variant of Orthogonal Matching Pursuit suitable for large linear systems, emphasizing its effectiveness for well-posed problems.

Contribution

It provides a convergence and numerical analysis of a low-memory implementation of the Self Projected Matching Pursuit algorithm, expanding its theoretical understanding.

Findings

Efficient low-memory implementation for large systems

Suitable for well-posed problems

Convergence properties established

Abstract

The convergence and numerical analysis of a low memory implementation of the Orthogonal Matching Pursuit greedy strategy, which is termed Self Projected Matching Pursuit, is presented. This approach renders an iterative way of solving the least squares problem with much less storage requirement than direct linear algebra techniques. Hence, it appropriate for solving large linear systems. The analysis highlights its suitability within the class of well posed problems.