Spectral-Informed Neural Networks Outperform Spectral Methods in High-dimensional PDEs

DOI10.48550/arXiv.2607.13566MaRDI QIDQ7284053arXivFDO

Authors Tianchi Yu, Ivan Oseledets

Publication date 15 July 2026

Published in arXiv

Tldr We develop a novel approach combining the strengths of physics-informed neural networks (PINNs) and spectral methods, by introducing coefficient decay scaling and basis embeddings motivated by harmonic analysis, achieving state-of-the-art results for steady and time-dependent PDEs in both middle- and high-dimensional settings.

Full work available at URL https://doi.org/10.48550/arXiv.2607.13566

arXiv classification

cs.AIcs.CEcs.LGmath.NA

This page was built for publication: Spectral-Informed Neural Networks Outperform Spectral Methods in High-dimensional PDEs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7284053)