A1052
Title: Structural inequalities in health across sub-Saharan Africa: A machine learning exploration of underlying determinants
Authors: Mercedes Tejeria-Martinez - University of Cantabria (Spain) [presenting]
Vanesa Jorda - University of Cantabria (Spain)
Jose Maria Sarabia - University of Cantabria (Spain)
Abstract: Health inequalities across sub-Saharan Africa (SSA), where disparities are exacerbated by limited access to healthcare and widespread socioeconomic inequities, represent critical barriers to equitable development. Advanced machine learning methodologies are leveraged to analyse health disparities in SSA countries using body mass index as a primary health indicator. Using demographic and health survey data, multiple machine learning algorithms are implemented to classify populations based on socioeconomic status, demographic profiles, and behavioural factors. The ensemble approach incorporates cross-validation model comparison techniques to optimise predictive accuracy and minimise algorithmic bias. The machine learning framework is designed to capture complex non-linear associations and interaction effects that conventional epidemiological approaches may miss. This computational approach aims to advance understanding of health inequality mechanisms in SSA, demonstrating how machine learning can enhance traditional public health research methodologies. The contribution to emerging digital health equity research is by establishing machine learning as a powerful tool for informing targeted, evidence-based interventions addressing systematic health disparities in developing regions.
