Titolo della tesi: Predictive Models for Public Health: Leveraging Machine Learning to Analyze Health Inequalities
Health inequalities remain a significant challenge in public health, influenced by complex interactions among socioeconomic, demographic, and environmental factors. This thesis explores the potential of machine learning techniques in modeling health disparities and predicting vulnerable population subgroups. Using data from the European Union Statistics on Income and Living Conditions (EU-SILC) 2009, this study develops predictive models to assess health status based on social determinants, including income, education, employment, housing conditions, and access to healthcare.
The research employs a cross-sectional observational design and integrates machine learning methodologies to enhance predictive accuracy beyond traditional statistical approaches. Key analytical steps include data preprocessing, feature selection, and model implementation using Google Collaboratory and Python, ensuring computational efficiency and methodological transparency.
Results indicate a strong correlation between socioeconomic conditions and health outcomes, with income, education level, and housing quality emerging as significant predictors of poor health. Machine learning models demonstrate superior predictive performance compared to conventional statistical methods, highlighting their utility in identifying at-risk groups and informing targeted health interventions.
Despite these advancements, the study acknowledges challenges related to data heterogeneity, model bias, and ethical considerations in AI-driven health analysis. The findings emphasize the necessity of integrating decision support systems (DSS) and predictive analytics into public health strategies to enhance resource allocation, reduce inequalities, and support evidence-based policymaking. This thesis contributes to the growing body of research advocating for data-driven approaches in epidemiology and health economics, reinforcing the role of machine learning in shaping a more equitable and sustainable healthcare system.