Learning the interaction rules of mobile animal groups with Graph Neural Networks
Supervisors:
Project Summary:
Collective movement is ubiquitous throughout the natural world; from fish schools, marching ants, to herds of wildebeest, living organisms live and move collectively. Understanding how animal groups are able to achieve large scale coordinated motion in the absence of centralised control is an important challenge with relevance for topics such as biodiversity conservation, swarm robotics, human crowd control, and cell migration. Recent developments in artificial intelligence have made it possible to develop and train intelligent algorithms that are able to process large complex datasets and extract meaningful insight from these data. A new class of machine learning (ML) models, graph neural networks (GNNs), present an exciting opportunity for understanding interacting systems including animal groups. These models can be trained on simulated or empirical data and can learn to predict the behaviour of complex interacting systems. This project seeks to develop a framework for using GNNs to model collective movement of animal groups across multiple scales. Specifically, the project will generate predictive GNN models of collect movement and train the models on data collected from two distinct biological systems, schooling fish and migratory wildebeest, in order to produce generalizable models for predicting animal movements in a range of systems.