Salvato in:
Dettagli Bibliografici
Autore principale: Chayti, El Mahdi
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2602.06824
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918501913985024
author Chayti, El Mahdi
author_facet Chayti, El Mahdi
contents Momentum methods, such as Polyak's Heavy Ball, are the standard for training deep networks but suffer from curvature-induced bias in stochastic settings, limiting convergence to suboptimal $\mathcal{O}(ε^{-4})$ rates. Existing corrections typically require expensive auxiliary sampling or restrictive smoothness assumptions. We propose \textbf{RanSOM}, a unified framework that eliminates this bias by replacing deterministic step sizes with randomized steps drawn from distributions with mean $η_t$. This modification allows us to leverage Stein-type identities to compute an exact, unbiased estimate of the momentum bias using a single Hessian-vector product computed jointly with the gradient, avoiding auxiliary queries. We instantiate this framework in two algorithms: \textbf{RanSOM-E} for unconstrained optimization (using exponentially distributed steps) and \textbf{RanSOM-B} for constrained optimization (using beta-distributed steps to strictly preserve feasibility). Theoretical analysis confirms that RanSOM recovers the optimal $\mathcal{O}(ε^{-3})$ convergence rate under standard bounded noise, and achieves optimal rates for heavy-tailed noise settings ($p \in (1, 2]$).
format Preprint
id arxiv_https___arxiv_org_abs_2602_06824
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle RanSOM: Second-Order Momentum with Randomized Scaling for Constrained and Unconstrained Optimization
Chayti, El Mahdi
Optimization and Control
Machine Learning
Momentum methods, such as Polyak's Heavy Ball, are the standard for training deep networks but suffer from curvature-induced bias in stochastic settings, limiting convergence to suboptimal $\mathcal{O}(ε^{-4})$ rates. Existing corrections typically require expensive auxiliary sampling or restrictive smoothness assumptions. We propose \textbf{RanSOM}, a unified framework that eliminates this bias by replacing deterministic step sizes with randomized steps drawn from distributions with mean $η_t$. This modification allows us to leverage Stein-type identities to compute an exact, unbiased estimate of the momentum bias using a single Hessian-vector product computed jointly with the gradient, avoiding auxiliary queries. We instantiate this framework in two algorithms: \textbf{RanSOM-E} for unconstrained optimization (using exponentially distributed steps) and \textbf{RanSOM-B} for constrained optimization (using beta-distributed steps to strictly preserve feasibility). Theoretical analysis confirms that RanSOM recovers the optimal $\mathcal{O}(ε^{-3})$ convergence rate under standard bounded noise, and achieves optimal rates for heavy-tailed noise settings ($p \in (1, 2]$).
title RanSOM: Second-Order Momentum with Randomized Scaling for Constrained and Unconstrained Optimization
topic Optimization and Control
Machine Learning
url https://arxiv.org/abs/2602.06824