Automated algorithm design for convex optimization problems with linear equality constraints

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ozaslan, Ibrahim K., Wu, Wuwei, Chen, Jie, Georgiou, Tryphon T., Jovanovic, Mihailo R.
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915513025691648
author Ozaslan, Ibrahim K.
Wu, Wuwei
Chen, Jie
Georgiou, Tryphon T.
Jovanovic, Mihailo R.
author_facet Ozaslan, Ibrahim K.
Wu, Wuwei
Chen, Jie
Georgiou, Tryphon T.
Jovanovic, Mihailo R.
contents Synthesis of optimization algorithms typically follows a {\em design-then-analyze\/} approach, which can obscure fundamental performance limits and hinder the systematic development of algorithms that operate near these limits. Recently, a framework grounded in robust control theory has emerged as a powerful tool for automating algorithm synthesis. By integrating design and analysis stages, fundamental performance bounds are revealed and synthesis of algorithms that achieve them is enabled. In this paper, we apply this framework to design algorithms for solving strongly convex optimization problems with linear equality constraints. Our approach yields a single-loop, gradient-based algorithm whose convergence rate is independent of the condition number of the constraint matrix. This improves upon the best known rate within the same algorithm class, which depends on the product of the condition numbers of the objective function and the constraint matrix.
format Preprint
id arxiv_https___arxiv_org_abs_2509_20746
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Automated algorithm design for convex optimization problems with linear equality constraints
Ozaslan, Ibrahim K.
Wu, Wuwei
Chen, Jie
Georgiou, Tryphon T.
Jovanovic, Mihailo R.
Optimization and Control
Systems and Control
Dynamical Systems
Synthesis of optimization algorithms typically follows a {\em design-then-analyze\/} approach, which can obscure fundamental performance limits and hinder the systematic development of algorithms that operate near these limits. Recently, a framework grounded in robust control theory has emerged as a powerful tool for automating algorithm synthesis. By integrating design and analysis stages, fundamental performance bounds are revealed and synthesis of algorithms that achieve them is enabled. In this paper, we apply this framework to design algorithms for solving strongly convex optimization problems with linear equality constraints. Our approach yields a single-loop, gradient-based algorithm whose convergence rate is independent of the condition number of the constraint matrix. This improves upon the best known rate within the same algorithm class, which depends on the product of the condition numbers of the objective function and the constraint matrix.
title Automated algorithm design for convex optimization problems with linear equality constraints
topic Optimization and Control
Systems and Control
Dynamical Systems
url https://arxiv.org/abs/2509.20746