Online estimation of the inverse of the Hessian for stochastic optimization with application to universal stochastic Newton algorithms - Archive ouverte HAL Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2024

Online estimation of the inverse of the Hessian for stochastic optimization with application to universal stochastic Newton algorithms

Résumé

This paper addresses second-order stochastic optimization for estimating the minimizer of a convex function written as an expectation. A direct recursive estimation technique for the inverse Hessian matrix using a Robbins-Monro procedure is introduced. This approach enables to drastically reduces computational complexity. Above all, it allows to develop universal stochastic Newton methods and investigate the asymptotic efficiency of the proposed approach. This work so expands the application scope of secondorder algorithms in stochastic optimization.
Fichier principal
Vignette du fichier
Newton_General.pdf (1.36 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04391570 , version 1 (12-01-2024)

Identifiants

Citer

Antoine Godichon-Baggioni, Wei Lu, Bruno Portier. Online estimation of the inverse of the Hessian for stochastic optimization with application to universal stochastic Newton algorithms. 2024. ⟨hal-04391570⟩
28 Consultations
26 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More