A phase transition in sampling from Restricted Boltzmann Machines

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kwon, Youngwoo, Qin, Qian, Wang, Guanyang, Wei, Yuchen
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914969951404032
author Kwon, Youngwoo
Qin, Qian
Wang, Guanyang
Wei, Yuchen
author_facet Kwon, Youngwoo
Qin, Qian
Wang, Guanyang
Wei, Yuchen
contents Restricted Boltzmann Machines are a class of undirected graphical models that play a key role in deep learning and unsupervised learning. In this study, we prove a phase transition phenomenon in the mixing time of the Gibbs sampler for a one-parameter Restricted Boltzmann Machine. Specifically, the mixing time varies logarithmically, polynomially, and exponentially with the number of vertices depending on whether the parameter $c$ is above, equal to, or below a critical value $c_\star\approx-5.87$. A key insight from our analysis is the link between the Gibbs sampler and a dynamical system, which we utilize to quantify the former based on the behavior of the latter. To study the critical case $c= c_\star$, we develop a new isoperimetric inequality for the sampler's stationary distribution by showing that the distribution is nearly log-concave.
format Preprint
id arxiv_https___arxiv_org_abs_2410_08423
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A phase transition in sampling from Restricted Boltzmann Machines
Kwon, Youngwoo
Qin, Qian
Wang, Guanyang
Wei, Yuchen
Machine Learning
Statistical Mechanics
Mathematical Physics
Probability
Computation
Restricted Boltzmann Machines are a class of undirected graphical models that play a key role in deep learning and unsupervised learning. In this study, we prove a phase transition phenomenon in the mixing time of the Gibbs sampler for a one-parameter Restricted Boltzmann Machine. Specifically, the mixing time varies logarithmically, polynomially, and exponentially with the number of vertices depending on whether the parameter $c$ is above, equal to, or below a critical value $c_\star\approx-5.87$. A key insight from our analysis is the link between the Gibbs sampler and a dynamical system, which we utilize to quantify the former based on the behavior of the latter. To study the critical case $c= c_\star$, we develop a new isoperimetric inequality for the sampler's stationary distribution by showing that the distribution is nearly log-concave.
title A phase transition in sampling from Restricted Boltzmann Machines
topic Machine Learning
Statistical Mechanics
Mathematical Physics
Probability
Computation
url https://arxiv.org/abs/2410.08423