Performance Comparison of Random Forest, SVM, and XGBoost Algorithms with SMOTE for Stunting Prediction

Stunting is a growth and development disorder caused by malnutrition, recurrent infections, and lack of psychosocial stimulation in which a child’s length or height is shorter than the growth standard for their age. With a prevalence of 21.5% in Indonesia by 2023, stunting is a global health problem that requires technology-based detection approaches for early intervention. This study evaluates the performance of three machine learning algorithms: Random Forest (RF), Support Vector Machine (SVM) and eXtreme Gradient Boosting (XGBoost) in predicting childhood stunting, and applying the SMOTE technique to handle data imbalance. The evaluation results show that XGBoost with SMOTE achieves the best performance with 87.83% accuracy, 85.75% precision, 91.59% recall, and 88.57% F1-score, superior to RF (84.56%) and SVM (68.59%). These results show that the combination of XGBoost and SMOTE is the best solution for an accurate and effective machine learning-based stunting detection system, so it can be used in public health programs to accelerate stunting risk identification.