Salvato in:
Dettagli Bibliografici
Autori principali: Roux, Christophe, Martínez-Rubio, David, Pokutta, Sebastian
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2501.18381
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866917906867027968
author Roux, Christophe
Martínez-Rubio, David
Pokutta, Sebastian
author_facet Roux, Christophe
Martínez-Rubio, David
Pokutta, Sebastian
contents We introduce a Riemannian optimistic online learning algorithm for Hadamard manifolds based on inexact implicit updates. Unlike prior work, our method can handle in-manifold constraints, and matches the best known regret bounds in the Euclidean setting with no dependence on geometric constants, like the minimum curvature. Building on this, we develop algorithms for g-convex, g-concave smooth min-max problems on Hadamard manifolds. Notably, one method nearly matches the gradient oracle complexity of the lower bound for Euclidean problems, for the first time.
format Preprint
id arxiv_https___arxiv_org_abs_2501_18381
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Implicit Riemannian Optimism with Applications to Min-Max Problems
Roux, Christophe
Martínez-Rubio, David
Pokutta, Sebastian
Optimization and Control
Machine Learning
We introduce a Riemannian optimistic online learning algorithm for Hadamard manifolds based on inexact implicit updates. Unlike prior work, our method can handle in-manifold constraints, and matches the best known regret bounds in the Euclidean setting with no dependence on geometric constants, like the minimum curvature. Building on this, we develop algorithms for g-convex, g-concave smooth min-max problems on Hadamard manifolds. Notably, one method nearly matches the gradient oracle complexity of the lower bound for Euclidean problems, for the first time.
title Implicit Riemannian Optimism with Applications to Min-Max Problems
topic Optimization and Control
Machine Learning
url https://arxiv.org/abs/2501.18381