Machine Learning Prediction Model for Early Student Academic Performance Evaluation in Video-Based Learning

The transformation of education norms from face-to-face teaching era to the Massive Open Online Courses (MOOCs) has created emergence educational technology domain for many students access e-learning platforms. However, there are some drawbacks especially in asynchronous video-based learning. A sense isolation could occur between teacher and if teachers do not interact much with Consequently, knowledge that is delivered by may reach effectively cause a drop student performance coming examination. Moreover, growth learning huge amount data on process video which provide boost mining research. Therefore, this research study aims introduce predictive model scrutinize number view based each chapter as well style, Felder-Silverman (FS) style deliver prediction individual early This tested different combination feature selection methods several handle imbalance such Synthetic Minority Oversampling Technique (SMOTE), SMOTE-TOMEK Adaptive (ADASYN) algorithms build machine compare performance. As result, proposed classifier Maximum Relevance Minimum Redundancy (MRMR) method SMOTE been achieved highest Area Under Curve (AUC) rate 0.93.