Mitigating spectral bias for the multiscale operator learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Xinliang, Xu, Bo, Cao, Shuhao, Zhang, Lei
Format: Preprint
Published: 2022
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916280405065728
author Liu, Xinliang
Xu, Bo
Cao, Shuhao
Zhang, Lei
author_facet Liu, Xinliang
Xu, Bo
Cao, Shuhao
Zhang, Lei
contents Neural operators have emerged as a powerful tool for learning the mapping between infinite-dimensional parameter and solution spaces of partial differential equations (PDEs). In this work, we focus on multiscale PDEs that have important applications such as reservoir modeling and turbulence prediction. We demonstrate that for such PDEs, the spectral bias towards low-frequency components presents a significant challenge for existing neural operators. To address this challenge, we propose a hierarchical attention neural operator (HANO) inspired by the hierarchical matrix approach. HANO features a scale-adaptive interaction range and self-attentions over a hierarchy of levels, enabling nested feature computation with controllable linear cost and encoding/decoding of multiscale solution space. We also incorporate an empirical $H^1$ loss function to enhance the learning of high-frequency components. Our numerical experiments demonstrate that HANO outperforms state-of-the-art (SOTA) methods for representative multiscale problems.
format Preprint
id arxiv_https___arxiv_org_abs_2210_10890
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Mitigating spectral bias for the multiscale operator learning
Liu, Xinliang
Xu, Bo
Cao, Shuhao
Zhang, Lei
Machine Learning
Artificial Intelligence
Numerical Analysis
Neural operators have emerged as a powerful tool for learning the mapping between infinite-dimensional parameter and solution spaces of partial differential equations (PDEs). In this work, we focus on multiscale PDEs that have important applications such as reservoir modeling and turbulence prediction. We demonstrate that for such PDEs, the spectral bias towards low-frequency components presents a significant challenge for existing neural operators. To address this challenge, we propose a hierarchical attention neural operator (HANO) inspired by the hierarchical matrix approach. HANO features a scale-adaptive interaction range and self-attentions over a hierarchy of levels, enabling nested feature computation with controllable linear cost and encoding/decoding of multiscale solution space. We also incorporate an empirical $H^1$ loss function to enhance the learning of high-frequency components. Our numerical experiments demonstrate that HANO outperforms state-of-the-art (SOTA) methods for representative multiscale problems.
title Mitigating spectral bias for the multiscale operator learning
topic Machine Learning
Artificial Intelligence
Numerical Analysis
url https://arxiv.org/abs/2210.10890