University of Tehran Press Interdisciplinary Journal of Management Studies 2981-0795 19 3 2026 05 31 Predicting University Entrance Examination Ranks by Developing a Stacking-Based Ensemble Machine Learning Algorithm 545 566 104297 10.22059/ijms.2025.385883.677186 EN Mohammad Reza Mehregan Department of Industrial Management, Faculty of Industrial Management and Technology, College of Management, University of Tehran, Tehran, Iran Arman Rezasoltani Department of Industrial Management, Faculty of Industrial Management and Technology, College of Management, University of Tehran, Tehran, Iran Amir Mohammad Khani Department of Industrial Management, Faculty of Industrial Management and Technology, College of Management, University of Tehran, Tehran, Iran Journal Article 2024 11 24 A key issue for planning and consulting is the accurate prediction of students’ rankings in important national university entrance exams, such as Iran’s nationwide university entrance examination, commonly known as the Konkur. Although machine learning has been increasingly used in educational data mining, most existing models have shown limited accuracy, are inadequately formulated, and lack sufficient optimization for practical application. This study introduces a novel stacking-based ensemble learning model that incorporates XGBoost, LightGBM, and CatBoost as base learners, with a linear regression model as a meta-learner to improve national rank prediction. The proposed model’s main hyperparameters were adjusted using the Optuna optimization framework to enhance the performance of each model. The model was trained and validated on a large dataset of over 73,000 student records from Ghalamchi Institute and evaluated using five-fold cross-validation with NRMSE and R² as performance measures. The results showed that the proposed model significantly outperformed baseline models, such as Random Forest, Gradient Boosting, and MLP Regressor, achieving NRMSE of 0.0659 and R² of 0.7735, which could be attributed to the effective integration of advanced learners with systematic hyperparameter optimization. This research provides a practical and scalable predictive tool that can support academic advisors, educators, and policymakers in making informed decisions, promoting equity in education, and guiding students through data-driven interventions. The use of stacking-based ensemble learning and automated hyperparameter optimization via Optuna distinguishes this study from prior research and is a meaningful step forward in the application of predictive analytics in high-risk educational settings. https://ijms.ut.ac.ir/article_104297_f59319e4a0067b07cfaf6680b0f3b2db.pdf