Epidemic spreading analysis through distributed simulation

Given the human and economical impacts of recent disease outbreaks, understanding epidemics has become a societal problem of prime importance. The increased mobility caused by the improvement of transportation has led some researchers to believe that a worldwide pandemic would occur before the end of the century. This thesis starts with a review of various epidemic models and their evolution. To obtain analytical results, these models have to resort to rather strong hypotheses, which are sometimes too unrealistic to be of any practical interest. To face this issue, a simulator was developed in collaboration with Real Impact Analytics. This program was designed to allow fine-grained modeling of each individual of the population through customizable behaviours. Using the simulator, we are able to confirm some theoretical results, and to study models for which analytical solutions are hard or impossible to obtain. Understanding the spreading dynamics of such models could be critical in the design of vaccination or quarantine strategies.