Bard College Professors Are Using NSF Grant For COVID-19 Forecasting Models
Professors at Bard College in Dutchess County have been awarded a National Science Foundation grant to fund a six-month study to develop forecasting models to better capture certain aspects of the COVID-19 pandemic.
Bard assistant professor of mathematics Matthew Junge is the lead investigator on the project. He says one advantage of a network model is that it is relatively easy to implement social distancing into the network.
“But we are trying to better understand how viruses or a virus like COVID-19,is spreading and how the measures that we’re we are taking to control it, impact its spread,” says Junge.
“I’d like to emphasize that this project has two parts. The first is about pure math, where we’ll hopefully be framing and answering interesting, new mathematical questions,” Junge says. “And the second part is taking those insights and applying them to the current pandemic.”
Junge further describes the aim of the $60,000 grant-funded project.
“We’re trying to develop a counterpoint to a lot of popular models that are out there. And most of the popular models are taking this very zoomed-out perspective on the disease and asking things about how an epidemic evolves in a whole country and, to do that, you have to make some assumptions about what happens on finer scales, like in communities,” Junge says. “And our perspective is more zoomed in. We want to look at what happens in communities and, to do that, we sort of sacrifice our accuracy about what you can say about the entire country of, the United States, for instance, but the benefit is we can answer more local questions, like should a specific community impose restaurant closures, for instance, or what should occupancy limits be on restaurants.”
He says bigger picture studies are already out there, which is why he saw a need to focus as he is.
“What’s really distinguishing it from a lot of the studies that are being quoted by the national press and various, the Administration is that we look at the local connections inside of communities, and those are usually ignored by bigger studies. In these zoomed-out models, they sort of picture cities and communities and schools as these black boxes where these diseases spread, but they don’t think about what happens at that level of detail,” Junge says. "But our study’s taking this opposite perspective of really finally modelling person-to-person connections that come up in our day-to-day lives, like who we socialize with, where we work, connections of that sort, and we ask how the disease spreads in this sort of zoomed-in picture.”
The grant was awarded through the NSF’s Rapid Response Research (RAPID) program, which provides support for urgent scientific research that responds to emergencies and unexpected events. Junge talks about how he came to propose the project.
“I’ve spent considerable time studying infection models. Recently, I’ve been looking at processes where people’s behavior actually changes as the epidemic evolves,” says Junge. “And having thought about those questions, that prepared me to propose and hopefully tackle this project for COVID-19.”
Bard College Professor Felicia Keesing brings biology to the project. She’s the David and Rosalie Rose Distinguished Professor of Science, Mathematics, and Computing. Grinnell College professor Nicole Eikmeier brings a computer science focus.