Collective behaviour in groups: flocks, schools and swarms

Large aggregations of living entities sometimes organise themselves into complex, moving patterns. Striking examples of this are bird flocks (most spectacularly the aerial display of huge numbers of starlings at dusk), fish schools and midge swarms. Collective behaviour also plays a key role on the microscopic scale of biological cells. In particular, in-vitro stem cells undergo complex dynamics as they evolve to form colonies and tissues; this process underlies future medical applications of stem cells for the controlled regeneration of biological tissue.

Perhaps surprisingly, this complex behaviour at the group level can emerge from simple actions of the individual members. Mathematical models show that a small set of rules, governing the interactions between individuals, can generate sophisticated, dynamic structures on large scales.

However, the inverse problem has not been tackled. That is, given a set of observations, can we infer the model parameters? A Bayesian approach to inference will be adopted, allowing flexible model structures and potential inclusion of expert information via the prior distribution.

This project will develop a model for such emergent collective behaviour, either directed to the macroscopic domain of birds or the microscopic realm of stem cells. Comparison to experimental observations will help deduce the biological, physical and geometrical processes which govern these dynamics, and can be expected to shed new light on collective behaviour in these systems.