Elliptic PDE learning is provably data-efficient

Nicolas Boullé, Diana Halikias, Alex Townsend
2/24/2023
cs.LGcs.AImath.NA

No official code found. Best implementation: NBoulle/pde-learning (7 stars, MIT, last commit 2024-11-20).

Abstract

PDE learning is an emerging field that combines physics and machine learning to recover unknown physical systems from experimental data. While deep learning models traditionally require copious amounts of training data, recent PDE learning techniques achieve spectacular results with limited data availability. Still, these results are empirical. Our work provides theoretical guarantees on the number of input-output training pairs required in PDE learning. Specifically, we exploit randomized numerical linear algebra and PDE theory to derive a provably data-efficient algorithm that recovers solution operators of 3D uniformly elliptic PDEs from input-output data and achieves an exponential convergence rate of the error with respect to the size of the training dataset with an exceptionally high probability of success.

View on arXivView PDF

Code Implementations(1)

NBoulle/pde-learningCommunity100%
MIT
70MATLAB, PythonFeb 22, 20231 years agoMIT

Similar Papers

Machine Learning for Partial Differential Equations

86% match1 repo1522023

Data-Efficient Operator Learning via Unsupervised Pretraining and In-Context Learning

84% match1 repo122024

Data-Efficient Operator Learning via Unsupervised Pretraining and In-Context Learning

84% match1 repo122026

PDE-DKL: PDE-constrained deep kernel learning in high dimensionality

84% match2025

Efficient Learning of PDEs via Taylor Expansion and Sparse Decomposition into Value and Fourier Domains

83% match2023

Solving Partial Differential Equations with Equivariant Extreme Learning Machines

83% match1 repo72024

Cite this paper

@article{boull2023elliptic,
  title  = {Elliptic PDE learning is provably data-efficient},
  author = {Nicolas Boullé and Diana Halikias and Alex Townsend},
  year   = {2023},
  eprint = {2302.12888},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2302.12888v2}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/2302.12888.svg)](https://codeofpaper.com/paper/2302.12888)

Discussion