Learning spatio-temporal patterns with Neural Cellular Automata

Alex D. Richardson*, Tibor Antal, Richard A. Blythe, Linus J. Schumacher

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract / Description of output

Neural Cellular Automata (NCA) are a powerful combination of machine learning and mechanistic modelling. We train NCA to learn complex dynamics from time series of images and PDE trajectories. Our method is designed to identify underlying local rules that govern large scale dynamic emergent behaviours. Previous work on NCA focuses on learning rules that give stationary emergent structures. We extend NCA to capture both transient and stable structures within the same system, as well as learning rules that capture the dynamics of Turing pattern formation in nonlinear Partial Differential Equations (PDEs). We demonstrate that NCA can generalise very well beyond their PDE training data, we show how to constrain NCA to respect given symmetries, and we explore the effects of associated hyperparameters on model performance and stability. Being able to learn arbitrary dynamics gives NCA great potential as a data driven modelling framework, especially for modelling biological pattern formation.
Original languageEnglish
Article numbere1011589
Pages (from-to)1-27
Number of pages27
JournalPLoS Computational Biology
Volume20
Issue number4
DOIs
Publication statusPublished - 26 Apr 2024

Keywords / Materials (for Non-textual outputs)

  • nlin.PS
  • cs.LG
  • cs.NE
  • math.DS
  • nlin.AO

Fingerprint

Dive into the research topics of 'Learning spatio-temporal patterns with Neural Cellular Automata'. Together they form a unique fingerprint.

Cite this