Spectral Informed Neural Network: An Efficient and Low-Memory PINN

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yu, Tianchi, Qi, Yiming, Oseledets, Ivan, Chen, Shiyi
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912673020510208
author Yu, Tianchi
Qi, Yiming
Oseledets, Ivan
Chen, Shiyi
author_facet Yu, Tianchi
Qi, Yiming
Oseledets, Ivan
Chen, Shiyi
contents With growing investigations into solving partial differential equations by physics-informed neural networks (PINNs), more accurate and efficient PINNs are required to meet the practical demands of scientific computing. One bottleneck of current PINNs is computing the high-order derivatives via automatic differentiation which often necessitates substantial computing resources. In this paper, we focus on removing the automatic differentiation of the spatial derivatives and propose a spectral-based neural network that substitutes the differential operator with a multiplication. Compared to the PINNs, our approach requires lower memory and shorter training time. Thanks to the exponential convergence of the spectral basis, our approach is more accurate. Moreover, to handle the different situations between physics domain and spectral domain, we provide two strategies to train networks by their spectral information. Through a series of comprehensive experiments, We validate the aforementioned merits of our proposed network.
format Preprint
id arxiv_https___arxiv_org_abs_2408_16414
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Spectral Informed Neural Network: An Efficient and Low-Memory PINN
Yu, Tianchi
Qi, Yiming
Oseledets, Ivan
Chen, Shiyi
Machine Learning
Artificial Intelligence
Numerical Analysis
Computational Physics
With growing investigations into solving partial differential equations by physics-informed neural networks (PINNs), more accurate and efficient PINNs are required to meet the practical demands of scientific computing. One bottleneck of current PINNs is computing the high-order derivatives via automatic differentiation which often necessitates substantial computing resources. In this paper, we focus on removing the automatic differentiation of the spatial derivatives and propose a spectral-based neural network that substitutes the differential operator with a multiplication. Compared to the PINNs, our approach requires lower memory and shorter training time. Thanks to the exponential convergence of the spectral basis, our approach is more accurate. Moreover, to handle the different situations between physics domain and spectral domain, we provide two strategies to train networks by their spectral information. Through a series of comprehensive experiments, We validate the aforementioned merits of our proposed network.
title Spectral Informed Neural Network: An Efficient and Low-Memory PINN
topic Machine Learning
Artificial Intelligence
Numerical Analysis
Computational Physics
url https://arxiv.org/abs/2408.16414