Discover physical concepts and equations with machine learning

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Li, Bao-Bing, Gu, Yi, Wu, Shao-Feng
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916702614192128
author Li, Bao-Bing
Gu, Yi
Wu, Shao-Feng
author_facet Li, Bao-Bing
Gu, Yi
Wu, Shao-Feng
contents Machine learning can uncover physical concepts or physical equations when prior knowledge from the other is available. However, these two aspects are often intertwined and cannot be discovered independently. We extend SciNet, which is a neural network architecture that simulates the human physical reasoning process for physics discovery, by proposing a model that combines Variational Autoencoders (VAE) with Neural Ordinary Differential Equations (Neural ODEs). This allows us to simultaneously discover physical concepts and governing equations from simulated experimental data across various physical systems. We apply the model to several examples inspired by the history of physics, including Copernicus' heliocentrism, Newton's law of gravity, Schrödinger's wave mechanics, and Pauli's spin-magnetic formulation. The results demonstrate that the correct physical theories can emerge in the neural network.
format Preprint
id arxiv_https___arxiv_org_abs_2412_12161
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Discover physical concepts and equations with machine learning
Li, Bao-Bing
Gu, Yi
Wu, Shao-Feng
Machine Learning
Disordered Systems and Neural Networks
Artificial Intelligence
Computational Physics
Machine learning can uncover physical concepts or physical equations when prior knowledge from the other is available. However, these two aspects are often intertwined and cannot be discovered independently. We extend SciNet, which is a neural network architecture that simulates the human physical reasoning process for physics discovery, by proposing a model that combines Variational Autoencoders (VAE) with Neural Ordinary Differential Equations (Neural ODEs). This allows us to simultaneously discover physical concepts and governing equations from simulated experimental data across various physical systems. We apply the model to several examples inspired by the history of physics, including Copernicus' heliocentrism, Newton's law of gravity, Schrödinger's wave mechanics, and Pauli's spin-magnetic formulation. The results demonstrate that the correct physical theories can emerge in the neural network.
title Discover physical concepts and equations with machine learning
topic Machine Learning
Disordered Systems and Neural Networks
Artificial Intelligence
Computational Physics
url https://arxiv.org/abs/2412.12161