Too Long; Didn't Read
Bayesian change point model to estimate the date that the number of new COVID-19 cases starts to flatten in different countries. The model was written in Pyro, a probabilistic programming language built on PyTorch. It is useful for estimating the date of change, not for predicting what will happen with the virus. It should not be mistaken for an amazing epidemiology model that will tell us when the quarantine will end, but instead a way of describing what we have already observed.