CHARLOTTESVILLE, Va. (WVIR) - Since the start of the outbreak, health organizations worldwide have been trying to predict the potential impact of COVID-19. Now, some students at the University of Virginia may be able to find a way to better predict the spread of the virus.
Models for the coronavirus released by the Virginia Department of Health can help people understand the number of lives that may have been saved by stay-at-home orders. One model may not be accurate for the entire state, according to applied mathematics professor Gianluca Guadagni at the University of Virginia.
"It’s very hard to get a very good model, and so what happens is that you have to cut corners. Because what we have right now, we don’t have detailed information,” Guadagni said.
The professor says models that show projections nationally - like those from the Centers for Disease Control and Prevention - can be helpful for those in the White House, but for smaller communities, like states, they may not be as helpful. That’s because their parameters don’t apply to every state.
"They can change based on something happening, and so keeping those parameters constant may damage the effectiveness of the model,” Guadagni explained. “A nice model is a model that has as few parameters as possible, but the price that you pay for that is that maybe the model is not going to describe very well what you want to describe”.
If you don’t have a mind that works with that math, Guadagni explains it using art: “Suppose that you want to paint something. If you have only two colors. You know you have some limits on it. If you have three colors, four colors, five colors, then you can get a better expression of what you want, right? And the same with parameters.”
Guadagni and applied mathematics students at UVA will now be able to create some new models based on research questions the students have related to COVID-19. That’s thanks in part to the $10,000 portion of the Jefferson Trust Grant he received. The project can help students understand that their learning is having real world impacts in real time, according to Guadagni.
“I want them to realize that there is something in it, and you can really model. So you can come up with some understanding just using the tools that we’re teaching in our courses,” Guadagni said.
The students participating in this project are still working on their research questions, so it will be some time before those models are made available to the public.