Exact Instance Compression for Convex Empirical Risk Minimization via Color Refinement

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Zhu, Bryan, Chen, Ziang
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866912864276578304
author Zhu, Bryan
Chen, Ziang
author_facet Zhu, Bryan
Chen, Ziang
contents Empirical risk minimization (ERM) can be computationally expensive, with standard solvers scaling poorly even in the convex setting. We propose a novel lossless compression framework for convex ERM based on color refinement, extending prior work from linear programs and convex quadratic programs to a broad class of differentiable convex optimization problems. We develop concrete algorithms for a range of models, including linear and polynomial regression, binary and multiclass logistic regression, regression with elastic-net regularization, and kernel methods such as kernel ridge regression and kernel logistic regression. Numerical experiments on representative datasets demonstrate the effectiveness of the proposed approach.
format Preprint
id arxiv_https___arxiv_org_abs_2602_00437
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Exact Instance Compression for Convex Empirical Risk Minimization via Color Refinement
Zhu, Bryan
Chen, Ziang
Optimization and Control
Machine Learning
Empirical risk minimization (ERM) can be computationally expensive, with standard solvers scaling poorly even in the convex setting. We propose a novel lossless compression framework for convex ERM based on color refinement, extending prior work from linear programs and convex quadratic programs to a broad class of differentiable convex optimization problems. We develop concrete algorithms for a range of models, including linear and polynomial regression, binary and multiclass logistic regression, regression with elastic-net regularization, and kernel methods such as kernel ridge regression and kernel logistic regression. Numerical experiments on representative datasets demonstrate the effectiveness of the proposed approach.
title Exact Instance Compression for Convex Empirical Risk Minimization via Color Refinement
topic Optimization and Control
Machine Learning
url https://arxiv.org/abs/2602.00437