Improving epidemic testing and containment strategies using machine learning

Abstract Containment of epidemic outbreaks entails great societal and economic costs. Cost-effective containment strategies rely on efficiently identifying infected individuals, making the best possible use available testing resources. Therefore, quickly optimal strategy is critical importance. Here, we demonstrate that machine learning can be used to identify which individuals are most beneficial test, automatically dynamically adapting characteristics disease outbreak. Specifically, simulate an outbreak using archetypal susceptible-infectious-recovered (SIR) model data about first confirmed cases train a neural network learns make predictions rest population. Using these predictions, manage contain more effectively than with standard approaches. Furthermore, how this method also when there possibility reinfection (SIRS model) eradicate endemic disease.