A short proof of near-linear convergence of adaptive gradient descent under fourth-order growth and convexity

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Davis, Damek, Drusvyatskiy, Dmitriy
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908965117362176
author Davis, Damek
Drusvyatskiy, Dmitriy
author_facet Davis, Damek
Drusvyatskiy, Dmitriy
contents Davis, Drusvyatskiy, and Jiang showed that gradient descent with an adaptive stepsize converges locally at a nearly-linear rate for smooth functions that grow at least quartically away from their minimizers. The argument is intricate, relying on monitoring the performance of the algorithm relative to a certain manifold of slow growth -- called the ravine. In this work, we provide a direct Lyapunov-based argument that bypasses these difficulties when the objective is in addition convex and a has a unique minimizer. As a byproduct of the argument, we obtain a more adaptive variant than the original algorithm with encouraging numerical performance.
format Preprint
id arxiv_https___arxiv_org_abs_2604_13393
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle A short proof of near-linear convergence of adaptive gradient descent under fourth-order growth and convexity
Davis, Damek
Drusvyatskiy, Dmitriy
Optimization and Control
Machine Learning
Davis, Drusvyatskiy, and Jiang showed that gradient descent with an adaptive stepsize converges locally at a nearly-linear rate for smooth functions that grow at least quartically away from their minimizers. The argument is intricate, relying on monitoring the performance of the algorithm relative to a certain manifold of slow growth -- called the ravine. In this work, we provide a direct Lyapunov-based argument that bypasses these difficulties when the objective is in addition convex and a has a unique minimizer. As a byproduct of the argument, we obtain a more adaptive variant than the original algorithm with encouraging numerical performance.
title A short proof of near-linear convergence of adaptive gradient descent under fourth-order growth and convexity
topic Optimization and Control
Machine Learning
url https://arxiv.org/abs/2604.13393