Accurately Estimating Total COVID-19 Infections using Information Theory

One of the most significant challenges in the early combat against COVID-19 was the difficulty in estimating the true magnitude of infections. Unreported infections drove up disease spread in numerous regions, made it very hard to accurately estimate the infectivity of the pathogen, therewith hampering our ability to react effectively. Despite the use of surveillance-based methods such as serological studies, identifying the true magnitude is still challenging today. This paper proposes an information theoretic approach for accurately estimating the number of total infections. Our approach is built on top of Ordinary Differential Equations (ODE) based models, which are commonly used in epidemiology and for estimating such infections. We show how we can help such models to better compute the number of total infections and identify the parameterization by which we need the fewest bits to describe the observed dynamics of reported infections. Our experiments show that our approach leads to not only substantially better estimates of the number of total infections but also better forecasts of infections than standard model calibration based methods. We additionally show how our learned parameterization helps in modeling more accurate what-if scenarios with non-pharmaceutical interventions. Our results support earlier findings that most COVID-19 infections were unreported and non-pharmaceutical interventions indeed helped to mitigate the spread of the outbreak. Our approach provides a general method for improving epidemic modeling which is applicable broadly.