Computer Science > Machine Learning
[Submitted on 15 Oct 2022 (v1), revised 3 Nov 2022 (this version, v5), latest version 9 Oct 2024 (v9)]
Title:Well-definedness of Physical Law Learning: The Uniqueness Problem
View PDFAbstract:Physical law learning is the ambiguous attempt at automating the derivation of governing equations with the use of machine learning techniques. The current literature focuses however solely on the development of methods to achieve this goal, and a theoretical foundation is at present missing. This paper shall thus serve as a first step to build a comprehensive theoretical framework for learning physical laws, aiming to provide reliability to according algorithms. One key problem consists in the fact that the governing equations might not be uniquely determined by the given data. We will study this problem in the common situation that a physical law is described by an ordinary or partial differential equation. For various different classes of differential equations, we provide both necessary and sufficient conditions for a function to uniquely determine the differential equation which is governing the phenomenon. We then use our results to devise numerical algorithms to determine whether a function solves a differential equation uniquely. Finally, we provide extensive numerical experiments showing that our algorithms in combination with common approaches for learning physical laws indeed allow to guarantee that a unique governing differential equation is learnt, without assuming any knowledge about the function, thereby ensuring reliability.
Submission history
From: Philipp Scholl [view email][v1] Sat, 15 Oct 2022 17:32:49 UTC (518 KB)
[v2] Wed, 19 Oct 2022 16:33:21 UTC (518 KB)
[v3] Thu, 27 Oct 2022 13:57:05 UTC (517 KB)
[v4] Fri, 28 Oct 2022 15:52:37 UTC (517 KB)
[v5] Thu, 3 Nov 2022 17:52:36 UTC (182 KB)
[v6] Tue, 8 Nov 2022 13:15:48 UTC (186 KB)
[v7] Tue, 20 Dec 2022 10:15:46 UTC (186 KB)
[v8] Thu, 5 Jan 2023 14:20:51 UTC (186 KB)
[v9] Wed, 9 Oct 2024 15:27:08 UTC (201 KB)
Current browse context:
cs.LG
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
Connected Papers (What is Connected Papers?)
CORE Recommender (What is CORE?)
IArxiv Recommender
(What is IArxiv?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.