SVD-NO: Learning PDE Solution Operators with SVD Integral Kernels

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Koren, Noam, Mackenbach, Ralf J. J., van Sloun, Ruud J. G., Radinsky, Kira, Freedman, Daniel
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915615041650688
author Koren, Noam
Mackenbach, Ralf J. J.
van Sloun, Ruud J. G.
Radinsky, Kira
Freedman, Daniel
author_facet Koren, Noam
Mackenbach, Ralf J. J.
van Sloun, Ruud J. G.
Radinsky, Kira
Freedman, Daniel
contents Neural operators have emerged as a promising paradigm for learning solution operators of partial differential equa- tions (PDEs) directly from data. Existing methods, such as those based on Fourier or graph techniques, make strong as- sumptions about the structure of the kernel integral opera- tor, assumptions which may limit expressivity. We present SVD-NO, a neural operator that explicitly parameterizes the kernel by its singular-value decomposition (SVD) and then carries out the integral directly in the low-rank basis. Two lightweight networks learn the left and right singular func- tions, a diagonal parameter matrix learns the singular values, and a Gram-matrix regularizer enforces orthonormality. As SVD-NO approximates the full kernel, it obtains a high de- gree of expressivity. Furthermore, due to its low-rank struc- ture the computational complexity of applying the operator remains reasonable, leading to a practical system. In exten- sive evaluations on five diverse benchmark equations, SVD- NO achieves a new state of the art. In particular, SVD-NO provides greater performance gains on PDEs whose solutions are highly spatially variable. The code of this work is publicly available at https://github.com/2noamk/SVDNO.git.
format Preprint
id arxiv_https___arxiv_org_abs_2511_10025
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SVD-NO: Learning PDE Solution Operators with SVD Integral Kernels
Koren, Noam
Mackenbach, Ralf J. J.
van Sloun, Ruud J. G.
Radinsky, Kira
Freedman, Daniel
Machine Learning
Neural operators have emerged as a promising paradigm for learning solution operators of partial differential equa- tions (PDEs) directly from data. Existing methods, such as those based on Fourier or graph techniques, make strong as- sumptions about the structure of the kernel integral opera- tor, assumptions which may limit expressivity. We present SVD-NO, a neural operator that explicitly parameterizes the kernel by its singular-value decomposition (SVD) and then carries out the integral directly in the low-rank basis. Two lightweight networks learn the left and right singular func- tions, a diagonal parameter matrix learns the singular values, and a Gram-matrix regularizer enforces orthonormality. As SVD-NO approximates the full kernel, it obtains a high de- gree of expressivity. Furthermore, due to its low-rank struc- ture the computational complexity of applying the operator remains reasonable, leading to a practical system. In exten- sive evaluations on five diverse benchmark equations, SVD- NO achieves a new state of the art. In particular, SVD-NO provides greater performance gains on PDEs whose solutions are highly spatially variable. The code of this work is publicly available at https://github.com/2noamk/SVDNO.git.
title SVD-NO: Learning PDE Solution Operators with SVD Integral Kernels
topic Machine Learning
url https://arxiv.org/abs/2511.10025