Infectious diseases pose a perpetual threat across the globe, devastating communities, and straining public health resources to their limit. The ease and speed of modern communications and transportation networks means policy makers are often playing catch-up to nascent epidemics, formulating critical, yet hasty, responses with insufficient, possibly inaccurate, information. In light of these difficulties, it is crucial to first understand the causes of a disease, then to predict its course, and finally to develop ways of controlling it. Mathematical modeling provides a methodical, in silico solution to all of these challenges, as we explore in this work. We accomplish these tasks with the aid of a surrogate modeling technique known as sparse grid interpolation, which approximates dynamical systems using a compact polynomial representation.

Our contributions to the disease modeling community are encapsulated in the following endeavors. We first explore transmission and recovery mechanisms for disease eradication, identifying a relationship between the reproductive potential of a disease and the maximum allowable disease burden. We then conduct a comparative computational study to improve simulation fits to existing case data by exploiting the approximation properties of sparse grid interpolants both on the global and local levels. Finally, we solve a joint optimization problem of periodically selecting field sensors and deploying public health interventions to progressively enhance the understanding of a metapopulation-based infectious disease system using a robust model predictive control scheme.