Hamiltonian analysis-based prediction service (HAPS)

The COVID-19 emergency has brought the analysis and mathematical modeling of epidemics to the forefront. As models become more sophisticated, and more accurate at capturing disease dynamics, new techniques must be developed to bring the data together with the model to get accurate predictions of the spread of infections. Wagner Associates proposes the development of a Hamiltonian Analysis-based Prediction Service (HAPS) that is rooted in analysis of stochastic epidemic models by approximating them as a Hamiltonian system. Here, collected data about the disease is used to numerically compute solution trajectories that can be used to extrapolate future dynamics, and predict dynamic shifts.  This method promises to be fast, accurate, and robust even to applications with limited available data. Additionally, solution dynamics can be analyzed based on parameter studies to generate an optimal control strategy based upon these forecasts.