High-dimensional Asymptotics of Langevin Dynamics in Spiked Matrix Models

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Liang, Tengyuan, Sen, Subhabrata, Sur, Pragya
Format: Preprint
Publié: 2022
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866913246956486656
author Liang, Tengyuan
Sen, Subhabrata
Sur, Pragya
author_facet Liang, Tengyuan
Sen, Subhabrata
Sur, Pragya
contents We study Langevin dynamics for recovering the planted signal in the spiked matrix model. We provide a "path-wise" characterization of the overlap between the output of the Langevin algorithm and the planted signal. This overlap is characterized in terms of a self-consistent system of integro-differential equations, usually referred to as the Crisanti-Horner-Sommers-Cugliandolo-Kurchan (CHSCK) equations in the spin glass literature. As a second contribution, we derive an explicit formula for the limiting overlap in terms of the signal-to-noise ratio and the injected noise in the diffusion. This uncovers a sharp phase transition -- in one regime, the limiting overlap is strictly positive, while in the other, the injected noise overcomes the signal, and the limiting overlap is zero.
format Preprint
id arxiv_https___arxiv_org_abs_2204_04476
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle High-dimensional Asymptotics of Langevin Dynamics in Spiked Matrix Models
Liang, Tengyuan
Sen, Subhabrata
Sur, Pragya
Statistics Theory
Machine Learning
Probability
We study Langevin dynamics for recovering the planted signal in the spiked matrix model. We provide a "path-wise" characterization of the overlap between the output of the Langevin algorithm and the planted signal. This overlap is characterized in terms of a self-consistent system of integro-differential equations, usually referred to as the Crisanti-Horner-Sommers-Cugliandolo-Kurchan (CHSCK) equations in the spin glass literature. As a second contribution, we derive an explicit formula for the limiting overlap in terms of the signal-to-noise ratio and the injected noise in the diffusion. This uncovers a sharp phase transition -- in one regime, the limiting overlap is strictly positive, while in the other, the injected noise overcomes the signal, and the limiting overlap is zero.
title High-dimensional Asymptotics of Langevin Dynamics in Spiked Matrix Models
topic Statistics Theory
Machine Learning
Probability
url https://arxiv.org/abs/2204.04476