Advanced Gradient Boosting Techniques for Predicting Obesity Risk: A Comprehensive Machine Learning Approach

Obesity, a major global health concern, is linked to several life-threatening conditions such as cardiovascular diseases, diabetes, and cancer. Early detection of obesity is crucial for effective intervention and prevention strategies. In this study, we investigate the performance of advanced machine learning techniques, specifically gradient boosting models-XGBoost, LightGBM, and CatBoost-for predicting obesity risk levels using medical, demographic, and lifestyle data. Rigorous preprocessing, hyperparameter tuning, and cross-validation were employed to ensure robust model development. XGBoost emerged as the most accurate model with an outstanding validation accuracy of 91.67%, followed by LightGBM and CatBoost. Despite minor overfitting in LightGBM, all three models demonstrated exceptional generalization capabilities. The findings highlight the potential of gradient boosting models for enhancing early obesity detection and personalized risk management in healthcare settings. Future work could focus on expanding the dataset, integrating explainable AI for better model interpretation, and exploring hybrid or deep learning models to further improve predictive performance. These results provide a solid foundation for implementing machine learning-based systems in public health programs aimed at combating the obesity epidemic.