Rank Inspired Neural Network for solving linear partial differential equations

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Peng, Wentao, Huang, Yunqing, Yi, Nianyu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912443884634112
author Peng, Wentao
Huang, Yunqing
Yi, Nianyu
author_facet Peng, Wentao
Huang, Yunqing
Yi, Nianyu
contents This paper proposes a rank inspired neural network (RINN) to tackle the initialization sensitivity issue of physics informed extreme learning machines (PIELM) when numerically solving partial differential equations (PDEs). Unlike PIELM which randomly initializes the parameters of its hidden layers, RINN incorporates a preconditioning stage. In this stage, covariance-driven regularization is employed to optimize the orthogonality of the basis functions generated by the last hidden layer. The key innovation lies in minimizing the off-diagonal elements of the covariance matrix derived from the hidden-layer output. By doing so, pairwise orthogonality constraints across collocation points are enforced which effectively enhances both the numerical stability and the approximation ability of the optimized function space.The RINN algorithm unfolds in two sequential stages. First, it conducts a non-linear optimization process to orthogonalize the basis functions. Subsequently, it solves the PDE constraints using linear least-squares method. Extensive numerical experiments demonstrate that RINN significantly reduces performance variability due to parameter initialization compared to PIELM. Incorporating an early stopping mechanism based on PDE loss further improves stability, ensuring consistently high accuracy across diverse initialization settings.
format Preprint
id arxiv_https___arxiv_org_abs_2506_17654
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Rank Inspired Neural Network for solving linear partial differential equations
Peng, Wentao
Huang, Yunqing
Yi, Nianyu
Numerical Analysis
This paper proposes a rank inspired neural network (RINN) to tackle the initialization sensitivity issue of physics informed extreme learning machines (PIELM) when numerically solving partial differential equations (PDEs). Unlike PIELM which randomly initializes the parameters of its hidden layers, RINN incorporates a preconditioning stage. In this stage, covariance-driven regularization is employed to optimize the orthogonality of the basis functions generated by the last hidden layer. The key innovation lies in minimizing the off-diagonal elements of the covariance matrix derived from the hidden-layer output. By doing so, pairwise orthogonality constraints across collocation points are enforced which effectively enhances both the numerical stability and the approximation ability of the optimized function space.The RINN algorithm unfolds in two sequential stages. First, it conducts a non-linear optimization process to orthogonalize the basis functions. Subsequently, it solves the PDE constraints using linear least-squares method. Extensive numerical experiments demonstrate that RINN significantly reduces performance variability due to parameter initialization compared to PIELM. Incorporating an early stopping mechanism based on PDE loss further improves stability, ensuring consistently high accuracy across diverse initialization settings.
title Rank Inspired Neural Network for solving linear partial differential equations
topic Numerical Analysis
url https://arxiv.org/abs/2506.17654