Agent-based models - from bacterial biofilms to primates

Speaker: Brian Logan

Abstract: In this talk I will give a brief overview of agent-based modelling (ABM) with examples drawn from some of the work in the Agents Lab. I will briefly explain what ABM is, when it is most useful, and the types of analyses it can support. To illustrate the range of problems to which ABM can be applied, I will give examples from two previous projects: a large scale (distributed) simulation of upgregulation in bacterial biofilms involving hundreds of thousands to millions of relatively simple agents over the space of a few hours, and a simulation of group decision making in baboons, which involves a relatively small number (50) of much more complex agents over a seven month period. In addition to spanning the size/complexity range, these two simulations also usefully contrast the kinds of questions one can ask of an ABM, from qualitative validation of a model to fitting empirical data and sensitivity analysis.