Enregistré dans:
Détails bibliographiques
Auteurs principaux: Parletta, Daniela Angela, Paudice, Andrea, Salzo, Saverio
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:https://arxiv.org/abs/2410.00573
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866916695235362816
author Parletta, Daniela Angela
Paudice, Andrea
Salzo, Saverio
author_facet Parletta, Daniela Angela
Paudice, Andrea
Salzo, Saverio
contents In this paper, we provide novel optimal (or near optimal) convergence rates for a clipped version of the stochastic subgradient method. We consider nonsmooth convex problems over possibly unbounded domains, under heavy-tailed noise that possesses only the first $p$ moments for $p \in \left]1,2\right]$. For the last iterate, we establish convergence in expectation for the objective values with rates of order $(\log^{1/p} k)/k^{(p-1)/p}$ and $1/k^{(p-1)/p}$, for anytime and finite-horizon respectively. We also derive new convergence rates, in expectation and with high probability, for the objective values along the average iterates--improving existing results by a $\log^{(2p-1)/p} k$ factor. Those results are applied to the problem of supervised learning with kernels demonstrating the effectiveness of our theory. Finally, we give preliminary experiments.
format Preprint
id arxiv_https___arxiv_org_abs_2410_00573
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle An Improved Analysis of the Clipped Stochastic subGradient Method under Heavy-Tailed Noise
Parletta, Daniela Angela
Paudice, Andrea
Salzo, Saverio
Optimization and Control
In this paper, we provide novel optimal (or near optimal) convergence rates for a clipped version of the stochastic subgradient method. We consider nonsmooth convex problems over possibly unbounded domains, under heavy-tailed noise that possesses only the first $p$ moments for $p \in \left]1,2\right]$. For the last iterate, we establish convergence in expectation for the objective values with rates of order $(\log^{1/p} k)/k^{(p-1)/p}$ and $1/k^{(p-1)/p}$, for anytime and finite-horizon respectively. We also derive new convergence rates, in expectation and with high probability, for the objective values along the average iterates--improving existing results by a $\log^{(2p-1)/p} k$ factor. Those results are applied to the problem of supervised learning with kernels demonstrating the effectiveness of our theory. Finally, we give preliminary experiments.
title An Improved Analysis of the Clipped Stochastic subGradient Method under Heavy-Tailed Noise
topic Optimization and Control
url https://arxiv.org/abs/2410.00573