• epidemic control
• statistical physics
• covid-19
• contact tracing
• epidemic models
• risk assessment
• belief propagation
• Bayesian inference

Digital contact tracing is a powerful tool that has the potential to play an important role in the efforts for the containment of an epidemic, such as the COVID-19 one. In this thesis the application of real-time contact tracing measures, which use Bayesian inference methods connected to statistical physics, is explored. The main focus is a contact tracing measure based on the Belief Propagation algorithm, which computes an approximation (risk) of the marginals of the posterior distribution of the states of the individuals given the observations (test results). The risk for each individual is then used to isolate and test the ones with the highest probability of being infected; by adopting this measure when there are still few infected individuals in the system, it is potentially possible to completely contain the epidemic, or at least delay the outbreak.