Jacopo Bindi — Politecnico di Torino #
Predicting epidemic processes on networks with partial observation #
Predicting epidemic processes on networks is still a challenging problem. Indeed in most of the cases 
dealing with incomplete or noisy data about the epidemics is unavoidable. Previous works provided a Belief 
propagation algorithm to deal with inference problems for irreversible stochasti epidemic model (SIR). 
They derived equations which allow to compute posterior distributions of the time evolution of the state 
of each node given some observation. We use this BP algorithm in order to predict characteristic quantity 
for SIR epidemic process relying only on partial observation at an early time.