04 Robust Agent-Based and Network Infectious Disease Models
The “Robust IDM” project will build on the foundations of IDM by developing agent-based and network models. The goal is to develop, expand and refine the agent-based modeling framework, leading to families of models that depart from rigid assumptions like a well-mixed population as adopted in ODE models.
- To develop for a large scale initiative like MfPH, our agent-based models will follow templates that can share common features such as the underlying social network and transmission settings, and are extendible in many dimensions, as finer scale epidemiological data and new knowledge comes available. In addition to taking advantage of the intrinsic conceptual advantages of transparency, flexibility and scalability, we also develop agent-based methods to address the curse of dimensionality, by combining agent-based methods with a parallel development of network and ODE analytics which make certain kinds of assumptions that lead to dramatic shortcuts in computation time. A well-defined class of network models that can “simulate the agent-based simulations quickly and more accurately than ODE models will also be developed.
Leads: Thomas Hurd (McMaster University, Hamilton) and Ali Asgary (York University, Toronto)
Team Members: Jason Brown, Arthur Charpentier, Helene Guerin, Jane Heffernan, Jeanette Jansen, Nathaniel Osgood, Sanjeev Seahra, Chris Soteros, James Watmough, Michael Wolfson