Analyzing Machine Learning Approaches for Diabetes Risk Prediction: Comparative Performance Assessment Using BRFSS Data

Diabetes represents a pressing global health concern, demanding effective predictive methodologies to alleviate its pervasive impact. This study conducts an exhaustive comparative analysis of three prominent machine learning (ML) approaches—Gaussian Naive Bayes, Decision Tree, and Artificial Neural Network (ANN)—for diabetes risk prediction. Leveraging the comprehensive Behavioral Risk Factor Surveillance System (BRFSS) dataset, rich insights into the performance nuances of each model are unveiled. Particularly noteworthy is the emergence of the ANN model as the most promising, exhibiting a remarkable accuracy of $84.73 \%$, outshining the Gaussian Naive Bayes and Decision Tree models. These findings underscore the transformative potential of advanced ML techniques, particularly ANN, in revolutionizing diabetes risk prediction and shaping future healthcare paradigms. By delving into the intricate efficacy of various ML algorithms in diabetes risk assessment, this study offers invaluable contributions to the field. It serves as a cornerstone for the development of more precise diagnostic tools and tailored therapeutic interventions aimed at enhancing diabetes management and mitigating its profound societal burden.