<![CDATA[Obesity is a growing health problem that can be influenced by various factors, such as diet, physical activity, and anthropometric conditions. This study aimed to predict obesity levels based on lifestyle and anthropometric factors using machine learning methods, namely Logistic Regression and XGBoost. The dataset used was obtained from the Kaggle platform, with a total of 2,111 data points and 17 attributes. The research process began with data preprocessing (label encoding and standard scaler), train-test data split (80:20), and validation using 5-Fold Stratified Cross-Validation. The results showed that the Logistic Regression model achieved 87.47% accuracy, while the XGBoost model achieved 95.98% accuracy in the train-test split and 94.35% ± 1.67% in the 5-Fold Cross-Validation. Feature importance analysis showed that Weight, Height, family_history_with_overweight, FAVC, and FAF were the most dominant predictors. Based on this comparison, the XGBoost model showed better performance and is recommended for predicting obesity levels.]]>
Copyrights © 2026
