Leveraging Ensemble-Learning Techniques to Predict Student Academic Performance

The significant growth of the Massive Open Online Course (MOCC) over last decade has promoted the rise of the educational data mining era in online education domain. This situation has created an opportunity for an educator to utilize the available data from MOOCs to facilitate student learning and performance. Therefore, this research study aims to introduce three types of ensemble learning methods, which are stacking, boosting, and bagging, to predict student performance on the online learning data that contain 480 students with 17 features. These techniques have to include the hyperparameter method to select the optimal number of input parameter to build the ensemble learning model. As a result, the proposed stacking type ensemble classifier has shown the highest prediction accuracy of approximately 94% and Area Under the Curve (AUC) of approximately 0.92. Results by stacking ensemble classifier have outperformed other ensemble classifiers, bagging and boosting as well as base classifiers.