Characterization of partially observed epidemics through Bayesian inference: application to COVID-19.

摘要: We demonstrate a Bayesian method for the "real-time" characterization and forecasting of partially observed COVID-19 epidemic. Characterization is estimation infection spread parameters using daily counts symptomatic patients. The designed to help guide medical resource allocation in early epoch outbreak. problem posed as one inference solved Markov chain Monte Carlo technique. data used this study was sourced before arrival second wave July 2020. proposed modeling approach, when applied at country level, generally provides accurate forecasts regional, state level. epidemiological model detected flattening curve California, after public health measures were instituted. also different disease dynamics specific regions New Mexico.