Solving Poisson Equations using Neural Walk-on-Spheres

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Nam, Hong Chul, Berner, Julius, Anandkumar, Anima
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910473284222976
author Nam, Hong Chul
Berner, Julius
Anandkumar, Anima
author_facet Nam, Hong Chul
Berner, Julius
Anandkumar, Anima
contents We propose Neural Walk-on-Spheres (NWoS), a novel neural PDE solver for the efficient solution of high-dimensional Poisson equations. Leveraging stochastic representations and Walk-on-Spheres methods, we develop novel losses for neural networks based on the recursive solution of Poisson equations on spheres inside the domain. The resulting method is highly parallelizable and does not require spatial gradients for the loss. We provide a comprehensive comparison against competing methods based on PINNs, the Deep Ritz method, and (backward) stochastic differential equations. In several challenging, high-dimensional numerical examples, we demonstrate the superiority of NWoS in accuracy, speed, and computational costs. Compared to commonly used PINNs, our approach can reduce memory usage and errors by orders of magnitude. Furthermore, we apply NWoS to problems in PDE-constrained optimization and molecular dynamics to show its efficiency in practical applications.
format Preprint
id arxiv_https___arxiv_org_abs_2406_03494
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Solving Poisson Equations using Neural Walk-on-Spheres
Nam, Hong Chul
Berner, Julius
Anandkumar, Anima
Machine Learning
Numerical Analysis
We propose Neural Walk-on-Spheres (NWoS), a novel neural PDE solver for the efficient solution of high-dimensional Poisson equations. Leveraging stochastic representations and Walk-on-Spheres methods, we develop novel losses for neural networks based on the recursive solution of Poisson equations on spheres inside the domain. The resulting method is highly parallelizable and does not require spatial gradients for the loss. We provide a comprehensive comparison against competing methods based on PINNs, the Deep Ritz method, and (backward) stochastic differential equations. In several challenging, high-dimensional numerical examples, we demonstrate the superiority of NWoS in accuracy, speed, and computational costs. Compared to commonly used PINNs, our approach can reduce memory usage and errors by orders of magnitude. Furthermore, we apply NWoS to problems in PDE-constrained optimization and molecular dynamics to show its efficiency in practical applications.
title Solving Poisson Equations using Neural Walk-on-Spheres
topic Machine Learning
Numerical Analysis
url https://arxiv.org/abs/2406.03494