Variational Neural Annealing

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hibat-Allah, Mohamed, Inack, Estelle M., Wiersema, Roeland, Melko, Roger G., Carrasquilla, Juan
Format: Preprint
Published: 2021
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929207845584896
author Hibat-Allah, Mohamed
Inack, Estelle M.
Wiersema, Roeland
Melko, Roger G.
Carrasquilla, Juan
author_facet Hibat-Allah, Mohamed
Inack, Estelle M.
Wiersema, Roeland
Melko, Roger G.
Carrasquilla, Juan
contents Many important challenges in science and technology can be cast as optimization problems. When viewed in a statistical physics framework, these can be tackled by simulated annealing, where a gradual cooling procedure helps search for groundstate solutions of a target Hamiltonian. While powerful, simulated annealing is known to have prohibitively slow sampling dynamics when the optimization landscape is rough or glassy. Here we show that by generalizing the target distribution with a parameterized model, an analogous annealing framework based on the variational principle can be used to search for groundstate solutions. Modern autoregressive models such as recurrent neural networks provide ideal parameterizations since they can be exactly sampled without slow dynamics even when the model encodes a rough landscape. We implement this procedure in the classical and quantum settings on several prototypical spin glass Hamiltonians, and find that it significantly outperforms traditional simulated annealing in the asymptotic limit, illustrating the potential power of this yet unexplored route to optimization.
format Preprint
id arxiv_https___arxiv_org_abs_2101_10154
institution arXiv
publishDate 2021
record_format arxiv
spellingShingle Variational Neural Annealing
Hibat-Allah, Mohamed
Inack, Estelle M.
Wiersema, Roeland
Melko, Roger G.
Carrasquilla, Juan
Disordered Systems and Neural Networks
Statistical Mechanics
Machine Learning
Quantum Physics
Many important challenges in science and technology can be cast as optimization problems. When viewed in a statistical physics framework, these can be tackled by simulated annealing, where a gradual cooling procedure helps search for groundstate solutions of a target Hamiltonian. While powerful, simulated annealing is known to have prohibitively slow sampling dynamics when the optimization landscape is rough or glassy. Here we show that by generalizing the target distribution with a parameterized model, an analogous annealing framework based on the variational principle can be used to search for groundstate solutions. Modern autoregressive models such as recurrent neural networks provide ideal parameterizations since they can be exactly sampled without slow dynamics even when the model encodes a rough landscape. We implement this procedure in the classical and quantum settings on several prototypical spin glass Hamiltonians, and find that it significantly outperforms traditional simulated annealing in the asymptotic limit, illustrating the potential power of this yet unexplored route to optimization.
title Variational Neural Annealing
topic Disordered Systems and Neural Networks
Statistical Mechanics
Machine Learning
Quantum Physics
url https://arxiv.org/abs/2101.10154