English

PDE-LEARN: Using Deep Learning to Discover Partial Differential Equations from Noisy, Limited Data

Machine Learning 2023-02-13 v2

Abstract

In this paper, we introduce PDE-LEARN, a novel deep learning algorithm that can identify governing partial differential equations (PDEs) directly from noisy, limited measurements of a physical system of interest. PDE-LEARN uses a Rational Neural Network, UU, to approximate the system response function and a sparse, trainable vector, ξ\xi, to characterize the hidden PDE that the system response function satisfies. Our approach couples the training of UU and ξ\xi using a loss function that (1) makes UU approximate the system response function, (2) encapsulates the fact that UU satisfies a hidden PDE that ξ\xi characterizes, and (3) promotes sparsity in ξ\xi using ideas from iteratively reweighted least-squares. Further, PDE-LEARN can simultaneously learn from several data sets, allowing it to incorporate results from multiple experiments. This approach yields a robust algorithm to discover PDEs directly from realistic scientific data. We demonstrate the efficacy of PDE-LEARN by identifying several PDEs from noisy and limited measurements.

Keywords

Cite

@article{arxiv.2212.04971,
  title  = {PDE-LEARN: Using Deep Learning to Discover Partial Differential Equations from Noisy, Limited Data},
  author = {Robert Stephany and Christopher Earls},
  journal= {arXiv preprint arXiv:2212.04971},
  year   = {2023}
}

Comments

25 pages, 7 figures, 9 tables

R2 v1 2026-06-28T07:28:06.503Z