We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Numerical Analysis

Title: Noise-aware Physics-informed Machine Learning for Robust PDE Discovery

Abstract: This work is concerned with discovering the governing partial differential equation (PDE) of a physical system. Existing methods have demonstrated the PDE identification from finite observations but failed to maintain satisfying results against noisy data, partly owing to suboptimal estimated derivatives and found PDE coefficients. We address the issues by introducing a noise-aware physics-informed machine learning (nPIML) framework to discover the governing PDE from data following arbitrary distributions. We propose training a couple of neural networks, namely solver and preselector, in a multi-task learning paradigm, which yields important scores of basis candidates that constitute the hidden physical constraint. After they are jointly trained, the solver network estimates potential candidates, e.g., partial derivatives, for the sparse regression algorithm to initially unveil the most likely parsimonious PDE, decided according to the information criterion. We also propose the denoising physics-informed neural networks (dPINNs), based on Discrete Fourier Transform (DFT), to deliver a set of the optimal finetuned PDE coefficients respecting the noise-reduced variables. The denoising PINNs are structured into forefront projection networks and a PINN, by which the formerly learned solver initializes. Our extensive experiments on five canonical PDEs affirm that the proposed framework presents a robust and interpretable approach for PDE discovery, applicable to a wide range of systems, possibly complicated by noise.
Comments: 13 pages, 8 figures, v2, v3: corrected typos and author names, v4, v5: improved notations
Subjects: Numerical Analysis (math.NA); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Computational Physics (physics.comp-ph)
Cite as: arXiv:2206.12901 [math.NA]
  (or arXiv:2206.12901v5 [math.NA] for this version)

Submission history

From: Pongpisit Thanasutives [view email]
[v1] Sun, 26 Jun 2022 15:29:07 GMT (1861kb,D)
[v2] Thu, 30 Jun 2022 13:18:35 GMT (1579kb,D)
[v3] Mon, 4 Jul 2022 16:31:21 GMT (1580kb,D)
[v4] Mon, 11 Jul 2022 10:32:26 GMT (1580kb,D)
[v5] Thu, 4 Aug 2022 15:25:12 GMT (1715kb,D)

Link back to: arXiv, form interface, contact.