Abstract:
Many colleges are reopening amid Fall 2020 of the COVID-19 pandemic with extreme measures in place: testing, dedensification, building closures, among others. We develop an agent-based network model to test intervention effectiveness. Our focus is on small colleges, which in aggregate serve over one million U.S. students, and have not been considered in-depth by existing models. We will survey how COVID-19 predictions are made for large areas like countries and cities, then go into detail about the models that came out this summer for disease spread on college campuses. From there, we will describe our model and findings. One of the more striking findings suggests that building closures may have unintended negative consequences. This is part of a broader observation that how students conduct themselves will determine if they get to enjoy, albeit a bit differently, the benefits of college life, or pass another year learning from a screen in their bedroom. Preprint available at https://arxiv.org/abs/2008.09597. 

Bio:
Matt completed his PhD at the University of Washington and his postdoc at Duke University. Noteworthy accomplishments from that time include giving 10 seminar talks during a 40 day road trip at the end of grad school, and competing in storyslams during his postdoc. Matt's research takes a probabilistic approach to particle systems from physics and biology. This includes models for chemical reactions, species proliferation, and epidemic outbreaks.

Zoom Link Access:
This talk will be given via Zoom, and the link is emailed to the CAM Seminar listserv the week of the talk. If you are not on the listserv, please contact Erika Fowler-Decatur to request the link.