Convergence of ADAM for Lipschitz Objective Functions

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ferrera, Juan, Gil, Javier Gómez
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914713264193536
author Ferrera, Juan
Gil, Javier Gómez
author_facet Ferrera, Juan
Gil, Javier Gómez
contents The aim of this paper is to prove the exponential convergence, local and global, of Adam algorithm under precise conditions on the parameters, when the objective function lacks differentiability. More precisely, we require Lipschitz continuity, and control on the gradient whenever it exists. We provide also examples of interesting functions that satisfies the required restrictions.
format Preprint
id arxiv_https___arxiv_org_abs_2403_08470
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Convergence of ADAM for Lipschitz Objective Functions
Ferrera, Juan
Gil, Javier Gómez
Optimization and Control
Classical Analysis and ODEs
Dynamical Systems
49J52 (Primary) 37N40, 46N10 (Secondary)
The aim of this paper is to prove the exponential convergence, local and global, of Adam algorithm under precise conditions on the parameters, when the objective function lacks differentiability. More precisely, we require Lipschitz continuity, and control on the gradient whenever it exists. We provide also examples of interesting functions that satisfies the required restrictions.
title Convergence of ADAM for Lipschitz Objective Functions
topic Optimization and Control
Classical Analysis and ODEs
Dynamical Systems
49J52 (Primary) 37N40, 46N10 (Secondary)
url https://arxiv.org/abs/2403.08470