Decomposing heterogeneous dynamical systems with graph neural networks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Allier, Cédric, Schneider, Magdalena C., Innerberger, Michael, Heinrich, Larissa, Bogovic, John A., Saalfeld, Stephan
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912560643571712
author Allier, Cédric
Schneider, Magdalena C.
Innerberger, Michael
Heinrich, Larissa
Bogovic, John A.
Saalfeld, Stephan
author_facet Allier, Cédric
Schneider, Magdalena C.
Innerberger, Michael
Heinrich, Larissa
Bogovic, John A.
Saalfeld, Stephan
contents Natural physical, chemical, and biological dynamical systems are often complex, with heterogeneous components interacting in diverse ways. We show how simple graph neural networks can be designed to jointly learn the interaction rules and the latent heterogeneity from observable dynamics. The learned latent heterogeneity and dynamics can be used to virtually decompose the complex system which is necessary to infer and parameterize the underlying governing equations. We tested the approach with simulation experiments of interacting moving particles, vector fields, and signaling networks. While our current aim is to better understand and validate the approach with simulated data, we anticipate it to become a generally applicable tool to uncover the governing rules underlying complex dynamics observed in nature.
format Preprint
id arxiv_https___arxiv_org_abs_2407_19160
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Decomposing heterogeneous dynamical systems with graph neural networks
Allier, Cédric
Schneider, Magdalena C.
Innerberger, Michael
Heinrich, Larissa
Bogovic, John A.
Saalfeld, Stephan
Machine Learning
Artificial Intelligence
Dynamical Systems
Natural physical, chemical, and biological dynamical systems are often complex, with heterogeneous components interacting in diverse ways. We show how simple graph neural networks can be designed to jointly learn the interaction rules and the latent heterogeneity from observable dynamics. The learned latent heterogeneity and dynamics can be used to virtually decompose the complex system which is necessary to infer and parameterize the underlying governing equations. We tested the approach with simulation experiments of interacting moving particles, vector fields, and signaling networks. While our current aim is to better understand and validate the approach with simulated data, we anticipate it to become a generally applicable tool to uncover the governing rules underlying complex dynamics observed in nature.
title Decomposing heterogeneous dynamical systems with graph neural networks
topic Machine Learning
Artificial Intelligence
Dynamical Systems
url https://arxiv.org/abs/2407.19160