Robust error estimates of PINN in one-dimensional boundary value problems for linear elliptic equations

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yoo, Jihahm, Lee, Haesung
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914937513705472
author Yoo, Jihahm
Lee, Haesung
author_facet Yoo, Jihahm
Lee, Haesung
contents In this paper, we study physics-informed neural networks (PINN) to approximate solutions to one-dimensional boundary value problems for linear elliptic equations and establish robust error estimates of PINN regardless of the quantities of the coefficients. In particular, we rigorously demonstrate the existence and uniqueness of solutions using the Sobolev space theory based on a variational approach. Deriving $L^2$-contraction estimates, we show that the error, defined as the mean square of the differences between the true solution and our trial function at the sample points, is dominated by the training loss. Furthermore, we show that as the quantities of the coefficients for the differential equation increase, the error-to-loss ratio rapidly decreases. Our theoretical and experimental results confirm the robustness of the error regardless of the quantities of the coefficients.
format Preprint
id arxiv_https___arxiv_org_abs_2407_14051
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Robust error estimates of PINN in one-dimensional boundary value problems for linear elliptic equations
Yoo, Jihahm
Lee, Haesung
Numerical Analysis
Analysis of PDEs
Primary: 34B05, 35A15, Secondary: 68T07, 65L10
In this paper, we study physics-informed neural networks (PINN) to approximate solutions to one-dimensional boundary value problems for linear elliptic equations and establish robust error estimates of PINN regardless of the quantities of the coefficients. In particular, we rigorously demonstrate the existence and uniqueness of solutions using the Sobolev space theory based on a variational approach. Deriving $L^2$-contraction estimates, we show that the error, defined as the mean square of the differences between the true solution and our trial function at the sample points, is dominated by the training loss. Furthermore, we show that as the quantities of the coefficients for the differential equation increase, the error-to-loss ratio rapidly decreases. Our theoretical and experimental results confirm the robustness of the error regardless of the quantities of the coefficients.
title Robust error estimates of PINN in one-dimensional boundary value problems for linear elliptic equations
topic Numerical Analysis
Analysis of PDEs
Primary: 34B05, 35A15, Secondary: 68T07, 65L10
url https://arxiv.org/abs/2407.14051