COMPARATIVE ANALYSIS OF PREDICTION MODELS FOR DIABETES BASED ON ANTHROPOMETRIC INDICATORS

The prevalence of type 2 diabetes mellitus (T2DM) continues to increase worldwide, emphasizing the need for early detection tools that are accurate, accessible, and cost-effective. Anthropometric indicators—such as body mass index (BMI), waist circumference (WC), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR), and body fat percentage—are widely used predictors of metabolic risk. Numerous statistical and machine learning models have been developed to forecast diabetes risk using these indicators. This study provides a comparative analysis of traditional regression-based models and modern machine learning algorithms to determine their predictive performance using anthropometric data.