43

Adekunle Kolawole Ojo
Landmark University, Omu-Aran, kwara State. Nigeria Department of Computer Science

Adebiyi Marion
Landmark University, Omu-Aran, kwara State. Nigeria Department of Computer Science

The Sub-Saharan Africa region remains disproportionately affected by the HIV/AIDS epidemic, accounting for the majority of global cases. Accurate predictive models are essential for effective intervention, resource allocation, and policy development to mitigate its impact. This study focuses on the development of a predictive model for HIV prevalence and transmission dynamics in Sub-Saharan Africa, leveraging machine learning techniques and epidemiological data. The model incorporates demographic, socio-economic, and behavioral factors, alongside environmental and healthcare access variables, to predict HIV trends at both regional and national levels. Using supervised learning approaches, including Random Forests and Gradient Boosting, as well as deep learning methods, the model identifies key determinants of HIV spread and evaluates their relative contributions. Validation was performed using historical HIV surveillance data, achieving high predictive accuracy. The model also includes a scenario-based projection feature, allowing stakeholders to assess the potential impact of different intervention strategies, such as increased antiretroviral therapy (ART) coverage or enhanced education campaigns. Preliminary results indicate that socio-economic disparities and healthcare accessibility are significant predictors of HIV prevalence. The developed model offers a robust tool for policymakers and healthcare providers, enabling data-driven decision-making to combat the HIV epidemic and promote sustainable health outcomes in the region.