Orateur
Description
Staged alert systems have been successfully implemented to minimize socioeconomic losses while avoiding overwhelming healthcare systems. Optimizing such systems can be formulated as a challenging two-stage stochastic mixed-integer programming problem with a discontinuous recourse function, where decision variables reside in a discrete space. Traditional simulation-based optimization techniques often assume continuity and smoothness in objective functions, while metaheuristics lack optimality guarantees. We propose a novel approach combining Gaussian process-based Bayesian optimization with a methodology tailored to handle discontinuities and efficiently navigate discrete search spaces. We establish global optimality guarantees with convergence properties. The efficacy of our method is demonstrated by optimizing stage thresholds in a pandemic alert system, providing a principled framework for discrete stochastic optimization under discontinuity.
