Novel Insights in Spatial Epidemiology Utilizing Explainable AI (XAI) and Remote Sensing

The COVID-19 pandemic has affected many aspects of human life around the world, due to its tremendous outcomes on public health and socio-economic activities. Policy makers have tried develop efficient responses based technologies advanced control methodologies, limit wide spreading virus in urban areas. However, techniques such as social isolation lockdown are short-term solutions that minimize spread cities do not invert long-term issues derive from climate change, air pollution planning challenges enhance ability. Thus, it seems crucial understand what kind factors assist or prevent virus. Although AI frameworks a very predictive ability data-driven procedures, they often struggle identify strong correlations among multidimensional data provide robust explanations. In this paper, we propose fusion heterogeneous, spatio-temporal dataset combine eight European spanning 1 January 2020 31 December 2021 describe atmospheric, socio-economic, health, mobility environmental all related potential links with COVID-19. Remote sensing key solution monitor availability green spaces between study period. So, evaluate benefits NIR RED bands satellite images calculate NDVI locate percentage vegetation cover each city for week our 2-year study. This novel is evaluated by tree-based machine learning algorithm utilizes ensemble trained make predictions daily cases deaths. Comparisons other justify robustness regression metrics RMSE MAE. Furthermore, explainable SHAP LIME utilized positive negative influence global local level, respect model’s A variation SHAP, namely treeSHAP, fast accurate