Neural network approaches for variance reduction in fluctuation formulas

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Pavliotis, Grigorios, Spacek, Renato, Stoltz, Gabriel, Vaes, Urbain
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913617726668800
author Pavliotis, Grigorios
Spacek, Renato
Stoltz, Gabriel
Vaes, Urbain
author_facet Pavliotis, Grigorios
Spacek, Renato
Stoltz, Gabriel
Vaes, Urbain
contents We propose a method utilizing physics-informed neural networks (PINNs) to solve Poisson equations that serve as control variates in the computation of transport coefficients via fluctuation formulas, such as the Green--Kubo and generalized Einstein-like formulas. By leveraging approximate solutions to the Poisson equation constructed through neural networks, our approach significantly reduces the variance of the estimator at hand. We provide an extensive numerical analysis of the estimators and detail a methodology for training neural networks to solve these Poisson equations. The approximate solutions are then incorporated into Monte Carlo simulations as effective control variates, demonstrating the suitability of the method for moderately high-dimensional problems where fully deterministic solutions are computationally infeasible.
format Preprint
id arxiv_https___arxiv_org_abs_2410_00278
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Neural network approaches for variance reduction in fluctuation formulas
Pavliotis, Grigorios
Spacek, Renato
Stoltz, Gabriel
Vaes, Urbain
Numerical Analysis
We propose a method utilizing physics-informed neural networks (PINNs) to solve Poisson equations that serve as control variates in the computation of transport coefficients via fluctuation formulas, such as the Green--Kubo and generalized Einstein-like formulas. By leveraging approximate solutions to the Poisson equation constructed through neural networks, our approach significantly reduces the variance of the estimator at hand. We provide an extensive numerical analysis of the estimators and detail a methodology for training neural networks to solve these Poisson equations. The approximate solutions are then incorporated into Monte Carlo simulations as effective control variates, demonstrating the suitability of the method for moderately high-dimensional problems where fully deterministic solutions are computationally infeasible.
title Neural network approaches for variance reduction in fluctuation formulas
topic Numerical Analysis
url https://arxiv.org/abs/2410.00278