Stochastic Differential Equations models for Least-Squares Stochastic Gradient Descent

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Schertzer, Adrien, Pillaud-Vivien, Loucas
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866914855998455808
author Schertzer, Adrien
Pillaud-Vivien, Loucas
author_facet Schertzer, Adrien
Pillaud-Vivien, Loucas
contents We study the dynamics of a continuous-time model of the Stochastic Gradient Descent (SGD) for the least-square problem. Indeed, pursuing the work of Li et al. (2019), we analyze Stochastic Differential Equations (SDEs) that model SGD either in the case of the training loss (finite samples) or the population one (online setting). A key qualitative feature of the dynamics is the existence of a perfect interpolator of the data, irrespective of the sample size. In both scenarios, we provide precise, non-asymptotic rates of convergence to the (possibly degenerate) stationary distribution. Additionally, we describe this asymptotic distribution, offering estimates of its mean, deviations from it, and a proof of the emergence of heavy-tails related to the step-size magnitude. Numerical simulations supporting our findings are also presented.
format Preprint
id arxiv_https___arxiv_org_abs_2407_02322
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Stochastic Differential Equations models for Least-Squares Stochastic Gradient Descent
Schertzer, Adrien
Pillaud-Vivien, Loucas
Machine Learning
Probability
We study the dynamics of a continuous-time model of the Stochastic Gradient Descent (SGD) for the least-square problem. Indeed, pursuing the work of Li et al. (2019), we analyze Stochastic Differential Equations (SDEs) that model SGD either in the case of the training loss (finite samples) or the population one (online setting). A key qualitative feature of the dynamics is the existence of a perfect interpolator of the data, irrespective of the sample size. In both scenarios, we provide precise, non-asymptotic rates of convergence to the (possibly degenerate) stationary distribution. Additionally, we describe this asymptotic distribution, offering estimates of its mean, deviations from it, and a proof of the emergence of heavy-tails related to the step-size magnitude. Numerical simulations supporting our findings are also presented.
title Stochastic Differential Equations models for Least-Squares Stochastic Gradient Descent
topic Machine Learning
Probability
url https://arxiv.org/abs/2407.02322