Extracting the governing equations of epidemics through sparse regression

In this work, our proposal is to extract the governing equations describing the time evolution of a given population where an epidemic spreading occurs, and for that end we employ a sparse regression model known as Sparse Identification of Nonlinear Dynamical Systems (SINDy). The equations resulting from such approach can then be used to make predictions about the future states of that group. The main advantage of SINDy is its interpretability, given its capabilities of recovering the equations from the system’s dynamics through a simple regression framework; and through them one can also acquire more information on the properties of such system.

Oct 16, 2023 — Oct 20, 2023

Conference on Complex Systems 2023