Leveraging Machine Learning to Unveil the Critical Role of Geographic Factors in COVID-19 Mortality in Mexico

Abstract

In this paper, we present an in-depthanalysis leveraging several renowned machine learningtechniques, including Snap Random Forest, XGBoost,Extra Trees, and Snap Decision Trees, to characterizecomorbidity factors influencing the Mexican population.Distinct from existing literature, our study undertakes acomprehensive exploration of algorithms within a definedsearch space, conducting experiments ranging fromcoarse to fine granularity. This approach, coupled withmachine learning-driven feature enhancement, enablesus to deeply characterize the factors most significantlyaffecting COVID-19 mortality rates within the Mexicandemographic. Contrary to other studies, which obscurethe identification of primary factors for local populations,our findings reveal that geographical factors such asresidence location hold greater significance than evencomorbidities, indicating that socioeconomic factors playa pivotal role in the survival outcomes of the Mexicanpopulation. This research not only contributes to thetargeted understanding of COVID-19 mortality driversin Mexico but also highlights the critical influence ofsocioeconomic determinants, offering valuable insightsfor public health strategies and policy formulation.