Salvato in:
Dettagli Bibliografici
Autori principali: Qin, Lei, Cantoni, Michael, Pu, Ye
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2510.20246
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913162079502336
author Qin, Lei
Cantoni, Michael
Pu, Ye
author_facet Qin, Lei
Cantoni, Michael
Pu, Ye
contents A variant of consensus based distributed gradient descent (\textbf{DGD}) is studied for finite sums of smooth but possibly non-convex functions. In particular, the local gradient term in the fixed step-size iteration of each agent is randomly perturbed to evade saddle points. Under regularity conditions, it is established that for sufficiently small step size and noise variance, each agent converges with high probability to a specified radius neighborhood of a common second-order stationary point, i.e., local minimizer. The rate of convergence is shown to be comparable to centralized first-order algorithms. Numerical experiments are presented to validate the efficacy of the proposed approach over standard \textbf{DGD} in a non-convex setting.
format Preprint
id arxiv_https___arxiv_org_abs_2510_20246
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Convergence Analysis of Noisy Distributed Gradient Descent for Non-convex Optimization -- Saddle Point Escape
Qin, Lei
Cantoni, Michael
Pu, Ye
Optimization and Control
A variant of consensus based distributed gradient descent (\textbf{DGD}) is studied for finite sums of smooth but possibly non-convex functions. In particular, the local gradient term in the fixed step-size iteration of each agent is randomly perturbed to evade saddle points. Under regularity conditions, it is established that for sufficiently small step size and noise variance, each agent converges with high probability to a specified radius neighborhood of a common second-order stationary point, i.e., local minimizer. The rate of convergence is shown to be comparable to centralized first-order algorithms. Numerical experiments are presented to validate the efficacy of the proposed approach over standard \textbf{DGD} in a non-convex setting.
title Convergence Analysis of Noisy Distributed Gradient Descent for Non-convex Optimization -- Saddle Point Escape
topic Optimization and Control
url https://arxiv.org/abs/2510.20246