Models of contagions can describe many network dynamics, such as cascading failures in economic systems, information diffusion, and pathogen transmission. 

with Dr. Jean-Gabriel Young
University of Vermont

Abstract: Models of contagions can describe many network dynamics, such as cascading failures in economic systems, information diffusion, and pathogen transmission. These models are particularly helpful when designing intervention---for example, deciding how to optimally roll out vaccines or designing robust economic systems. The idea is, roughly, to use realistic contagion models to test counterfactuals, and optimize the system's parameters to increase the likelihood of favorable outcomes. This is a difficult numerical optimization problem. A leading computational approach is message passing, which allows for the rapid and direct computation of the distribution over outcomes.  In this talk, I will give a pedagogical introduction to message passing. I will then highlight some well-known issues with this framework, chiefly that it overestimates the likelihood that a contagion will spread. I will then discuss two promising approaches for correcting these issues: the neighborhood message passing (NMP) framework and graph machine learning techniques.

Bio: Jean-Gabriel Young is an Assistant Professor in the Department of Mathematics and Statistics at the University of Vermont, and a faculty member of both the Vermont Complex Systems Institute and the Translational Global Infectious Diseases Research Center. His research focuses on statistical inference, epidemiology, and the structure and dynamics of complex systems. He was previously a James S. McDonnell Foundation Fellow at the University of Michigan’s Center for the Study of Complex Systems, working with Prof. Mark Newman. He earned his Ph.D. in Physics from Université Laval.