Dynamic message-passing approach to epidemic spreading

Abstract

In this work, we delve into the analysis of the propagation of infectious diseases described by the SIR epidemiological model on networks whose connections change over time. We use a fast dynamic message-passing method to estimate the marginal probabilities that each node in one of these networks is in a given state at a certain time. After testing the validity of the predictions given by this approach, we apply them to the problem of inferring the origin of an epidemic on a dynamic network.