Optimal disease outbreak decisions using stochastic simulation

Management policies for disease outbreaks balance the expected morbidity and mortality costs versus the cost of intervention policies. We present a methodology for dynamic determination of optimal policies in a stochastic compartmental model with parameter uncertainty. Our approach is to first carry out sequential Bayesian estimation of outbreak parameters and then solve the dynamic programming equations. The latter step is simulation-based and relies on regression Monte Carlo techniques. To improve performance we investigate lasso regression and global policy iteration. Comparisons demonstrate the realized cost savings of choosing interventions based on the computed dynamic policy over simpler decision rules.