A gradient descent algorithm built on approximate discrete gradients - Archive ouverte HAL
Communication Dans Un Congrès Année : 2022

A gradient descent algorithm built on approximate discrete gradients

Résumé

We propose an optimization method obtained by the approximation of a novel discretization approach for gradient dynamics recently proposed by the authors. It is shown that the proposed algorithm ensures convergence for all amplitudes of the step size, contrarily to classical implementations.
Fichier principal
Vignette du fichier
_Final__A_gradient_descent_algorithm_built_on_approximate_discrete_gradients.pdf (655.1 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04227956 , version 1 (04-10-2023)

Identifiants

Citer

Alessio Moreschini, Mattia Mattioni, Salvatore Monaco, Dorothée Normand-Cyrot. A gradient descent algorithm built on approximate discrete gradients. 26th International Conference on System Theory, Control and Computing (ICSTCC 2022), Oct 2022, Sinaia, Romania. pp.343-348, ⟨10.1109/ICSTCC55426.2022.9931872⟩. ⟨hal-04227956⟩
41 Consultations
320 Téléchargements

Altmetric

Partager

More