UArizona Applied Math COVID-19 Working Group

Data assimilation and parameter estimation with an eye towards COVID modeling
Paul Carter, Department of Mathematics, University of Minnesota

Abstract: Data assimilation is a useful framework for incorporating observations into dynamical models, in order to estimate and predict states, as well as estimate model parameters. I'll provide a (very) brief introduction to these ideas, with some examples in compartmental models of epidemiology.

Data-Driven Model Discovery and COVID-19
Bill Fries, Program in Applied Mathematics, University of Arizona

Abstract:  With the vast amounts of available data surrounding the spread of COVID-19, the use of data-motivated models can help give a potential look into what factors contribute most to its spread and what its long-term trajectory might look like. In this talk, I will outline SINDy (Sparse Identification of Nonlinear Dynamics) as a tool for model discovery and explore some potential applications in the COVID-19 era. Specifically, we will look at a simple compartmental model and SINDy's robustness, the potential uses of this tool in both a standard SIR model and one with spacial dependency which incorporates transportation data and considers travel between cities.

Zoom:  https://arizona.zoom.us/j/183045568