Solving Large Scale Quadratic Constrained Basis Pursuit

TL;DR

This paper introduces an efficient algorithm for large-scale quadratic constrained basis pursuit, significantly accelerating solution times compared to traditional interior point methods.

Contribution

It presents a novel algorithm inspired by ADMM and operator splitting, tailored for large-scale quadratic constrained basis pursuit problems.

Findings

Achieves 50-100x speedup over baseline interior point methods

Demonstrates effectiveness on large-scale problems

Validates efficiency through experimental results

Abstract

Inspired by alternating direction method of multipliers and the idea of operator splitting, we propose a efficient algorithm for solving large-scale quadratically constrained basis pursuit. Experimental results show that the proposed algorithm can achieve 50~~100 times speedup when compared with the baseline interior point algorithm implemented in CVX.