Improving student test score prediction results using Machine Learning with R support

Abstract

Predicting final exam scores of students is important to forecast student accomplishment in the learning process and functioned to help the students who might fail. This paper presents a method that based on machine learning techniques to enhance prediction results by handling the imbalanced data issues. The experimental results achieves the highest performance with 43.70% of Precision, 45.1% of Recall and 44.04% of F1.