Date of Award
2026
Degree Type
Open Access Dissertation
Degree Name
Information Systems and Technology, PhD
Program
Center for Information Systems and Technology
Advisor/Supervisor/Committee Chair
Itamar Shabtai
Dissertation or Thesis Committee Member
Warren Roberts
Dissertation or Thesis Committee Member
Zachary Dodds
Dissertation or Thesis Committee Member
Julie Medero & Conrad Shayo
Terms of Use & License Information
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Rights Information
© 2026 Maria Assumpta Komugabe
Keywords
Geographically Weighted Regression, GIS, Machine learning/GEOAI, Malaria Monitoring, Random Forest machine learning, Spatial Autoregressive Models
Subject Categories
Geographic Information Sciences | Public Health
Abstract
Malaria remains a critical public health challenge in Uganda—responsible for an estimated 13.2 million cases annually yet control efforts are continually undermined by climate variability and systemic supply chain failures. This study investigated the relationship between climatic variables (rainfall, temperature, humidity) and malaria incidence, evaluated the efficacy of the current supply management system, and assessed how AI-predictive models and GIS tools can optimize resource distribution. Employing a multi-scalar mixed-methods approach, the research utilized Spatial Autoregressive Models (SAR), Geographically Weighted Regression (GWR), and Random Forest machine learning (R2 = .86) to analyze transmission dynamics. Key findings reveal that minimum temperature is the strongest predictor of transmission (p < .001), while rainfall acts as a reliable 30-day (Lag-1) lead indicator. Emerging Hot Spot Analysis (EHSA) isolated 11 "Intensifying" districts requiring immediate saturation and 32 "Sporadic" zones driven by seasonal volatility. Furthermore, the evaluation of the current supply system exposed a "phantom mortality reduction," where rising stockouts (averaging 53 consecutive days in the Central Region) artificially lowered reported deaths. Significantly, a spatial spillover effect (rho = 0.27) confirmed that uncoordinated interventions fail as high-burden districts export risk to neighboring areas. This research provides three major advancements: methodologically, it combines spatial econometrics with machine learning for high-precision risk forecasting; practically, it introduces an AI-GIS dashboard and Early Warning System to enable a "Self-Healing Supply Chain" and theoretically, the research extends the Health System Resilience Theory established by Kruk et al. (2015) and Blanchet et al. (2017) by introducing the Adaptive Digital Resilience Model (ADRM). Despite these technical gains, the study notes that successful "last mile" delivery depends on closing the digital divide among Village Health Teams (VHTs) to protect the most vulnerable populations.
ISBN
9798244861044
Recommended Citation
Komugabe, Maria Assumpta. (2026). Integrating AI and GIS for Climate-Driven Malaria Monitoring and Demand-Based Resource Distribution and Supply Optimization in Low Developing Countries. CGU Theses & Dissertations, 1110. https://scholarship.claremont.edu/cgu_etd/1110.
