Article Dans Une Revue Journal of Statistical Mechanics: Theory and Experiment Année : 2024

High-dimensional non-convex landscapes and gradient descent dynamics

Résumé

In these lecture notes we present different methods and concepts developed in statistical physics to analyze gradient descent dynamics in high-dimensional non-convex landscapes. Our aim is to show how approaches developed in physics, mainly statistical physics of disordered systems, can be used to tackle open questions on high-dimensional dynamics in machine learning.

Fichier principal
Vignette du fichier
2308.03754v2.pdf (4.78 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-04937911 , version 1 (11-02-2025)

Licence

Identifiants

Citer

Tony Bonnaire, Davide Ghio, Kamesh Krishnamurthy, Francesca Mignacco, Atsushi Yamamura, et al.. High-dimensional non-convex landscapes and gradient descent dynamics. Journal of Statistical Mechanics: Theory and Experiment, 2024, 2024 (10), pp.104004. ⟨10.1088/1742-5468/ad2929⟩. ⟨hal-04937911⟩
107 Consultations
328 Téléchargements

Altmetric

Partager

  • More