Covid-19 Intervention Policy Optimization Using a Multi-population Evolutionary Algorithm

Abstract

The rapid spread of COVID-19, caused by the SARS-CoV-2 virus, has resulted in and continues to pose a significant threat to global health. We proposed a predictive model based on the gated recurrent unit (GRU) that investigates the influence of non-pharmaceutical interventions (NPIs) on the progression of COVID-19. The proposed model is validated by applying in different states in the United States, although it can be generalized to any region of interest in the world. The results show that the predictive model can achieve accurate forecasts across the entire US. The forecast is then utilized to identify the optimal mitigation policies. The goal is to find the best stringency level for each policy that can minimize the total number of new Covid-19 cases while minimizing the mitigation costs. A meta-heuristics method, named multi-population evolutionary algorithm with differential evolution (MPEA-DE), has been developed. The goal is to identify optimal mitigation strategies that minimize COVID-19 infection cases while controlling the costs and other negative impacts. We compared the optimal mitigation strategies identified by the MPEA-DE model with the random search and the blind greedy search strategies. The results show that MPEA-DE performs better than other baseline models based on prescription dominance.