PLANNING OF TEMPORARY HOSPITALS THROUGH OPTIMIZATION VIA SIMULATION: AN APPROACH BASED ON ADAPTIVE METAMODELING

Temporary hospitals are essential to alleviating healthcare system overload during emergencies such as pandemics and natural disasters. However, their planning requires effective strategies for resource allocation and operational optimization. This study presents Adaptive Metamodeling-Based Simulation Optimization (AMSO), an innovative approach that utilizes advanced techniques such as Bagging Gradient Boosted Trees, hyperparameter optimization, and metaheuristics. Applied to a temporary hospital in Brazil during COVID-19, AMSO outperformed traditional methods like EGO and GA, achieving results equivalent to GA with fewer experiments and reduced computational time. This efficiency highlights AMSO as a promising tool for optimized temporary hospital planning, minimizing patient length of stay and contributing to a more agile and effective healthcare management system.