Distributed Stochastic Zeroth-Order Optimization with Compressed Communication

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Hua, Youqing, Liu, Shuai, Hong, Yiguang, Ren, Wei
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866909793849966592
author Hua, Youqing
Liu, Shuai
Hong, Yiguang
Ren, Wei
author_facet Hua, Youqing
Liu, Shuai
Hong, Yiguang
Ren, Wei
contents The dual challenges of prohibitive communication overhead and the impracticality of gradient computation due to data privacy or black-box constraints in distributed systems motivate this work on communication-constrained gradient-free optimization. We propose a stochastic distributed zeroth-order algorithm (Com-DSZO) requiring only two function evaluations per iteration, integrated with general compression operators. Rigorous analysis establishes its sublinear convergence rate for both smooth and nonsmooth objectives, while explicitly elucidating the compression-convergence trade-off. Furthermore, we develop a variance-reduced variant (VR-Com-DSZO) under stochastic mini-batch feedback. The empirical algorithm performance are illustrated with numerical examples.
format Preprint
id arxiv_https___arxiv_org_abs_2503_17429
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Distributed Stochastic Zeroth-Order Optimization with Compressed Communication
Hua, Youqing
Liu, Shuai
Hong, Yiguang
Ren, Wei
Optimization and Control
Multiagent Systems
The dual challenges of prohibitive communication overhead and the impracticality of gradient computation due to data privacy or black-box constraints in distributed systems motivate this work on communication-constrained gradient-free optimization. We propose a stochastic distributed zeroth-order algorithm (Com-DSZO) requiring only two function evaluations per iteration, integrated with general compression operators. Rigorous analysis establishes its sublinear convergence rate for both smooth and nonsmooth objectives, while explicitly elucidating the compression-convergence trade-off. Furthermore, we develop a variance-reduced variant (VR-Com-DSZO) under stochastic mini-batch feedback. The empirical algorithm performance are illustrated with numerical examples.
title Distributed Stochastic Zeroth-Order Optimization with Compressed Communication
topic Optimization and Control
Multiagent Systems
url https://arxiv.org/abs/2503.17429