English

A Cyclic Coordinate Descent Method for Convex Optimization on Polytopes

Optimization and Control 2023-04-28 v2

Abstract

Coordinate descent algorithms are popular for huge-scale optimization problems due to their low cost per-iteration. Coordinate descent methods apply to problems where the constraint set is separable across coordinates. In this paper, we propose a new variant of the cyclic coordinate descent method that can handle polyhedral constraints provided that the polyhedral set does not have too many extreme points such as L1-ball and the standard simplex. Loosely speaking, our proposed algorithm PolyCD, can be viewed as a hybrid of cyclic coordinate descent and the Frank-Wolfe algorithms. We prove that PolyCD has a O(1/k) convergence rate for smooth convex objectives. Inspired by the away-step variant of Frank-Wolfe, we propose PolyCDwA, a variant of PolyCD with away steps which has a linear convergence rate when the loss function is smooth and strongly convex. Empirical studies demonstrate that PolyCDwA achieves strong computational performance for large-scale benchmark problems including L1-constrained linear regression, L1-constrained logistic regression and kernel density estimation.

Keywords

Cite

@article{arxiv.2303.07642,
  title  = {A Cyclic Coordinate Descent Method for Convex Optimization on Polytopes},
  author = {Rahul Mazumder and Haoyue Wang},
  journal= {arXiv preprint arXiv:2303.07642},
  year   = {2023}
}
R2 v1 2026-06-28T09:15:36.043Z