1 Department of Computer Science, University of Ghana, Ghana.
2 Department of Agriculture, University of Ilorin, Nigeria.
3 Western Michigan University, School of environment, geography and sustainability, USA.
Research Article
World Journal of Advanced Research and Reviews, 2024, 24(03), 3610-3618
Received on 02 November 2024; revised on 28 December 2024; accepted on 30 December 2024
Chronic diseases remain one of the leading causes of preventable morbidity in the United States, yet early detection at the community level is limited by delayed reporting and fragmented data sources. This study presents an integrated framework that combines geospatial analytics and predictive modeling to support early surveillance of chronic-disease risk across census tracts. Using publicly available datasets from the CDC PLACES project, the American Community Survey, EPA air-quality monitoring, and food-access indicators, we engineered spatial features including hotspot clusters, spatial-lag variables, and environmental exposure models. Logistic Regression, Random Forest, and XGBoost were trained to classify high-risk areas, with model performance evaluated using AUC, precision, recall, and geographically weighted diagnostics. Findings demonstrate that incorporating spatial dependencies significantly improves predictive accuracy and enhances the interpretability of risk patterns. The proposed framework can support public health agencies in proactively identifying emerging clusters, prioritizing resource allocation, and implementing timely community-level interventions.
Geospatial analytics; Predictive modeling; Chronic-disease surveillance; Spatial-lag features; Hotspot analysis; Machine learning
Preview Article PDF
Justine Aku Azigi, Abdullahi Abdulkareem and Kwadwo Frimpong. Integrating geospatial analytics and predictive modeling for early chronic-disease surveillance. World Journal of Advanced Research and Reviews, 2024, 24(3), 3610-3618. Article DOI: https://doi.org/10.30574/wjarr.2024.24.3.3446