Predicting student’s performance using machine learning methods: A systematic literature review

Student’s performance is a success factor in higher education institutions. The excellent record of academic achievements raises the institution’s ranking as one of the criteria for a high-quality university. Predicting and analyzing the performance of the student is essential to assist educators in identifying weaknesses and enhancing the academic scores. However, achieving accurate predictions is challenging due to huge amount of educational data. The main reason behind this, is the lack of research on exploring different prediction methods and key attributes that influence the student’s academic performance. Hence, this systematic review intends to explore the current machine learning methods and attributes used in predicting the student’s performance. Several online databases were used to perform a systematic search of data-driven studies. The analysis and assessment of 30 selected articles revealed five main prediction methods: artificial neural networks (ANNs), decision trees, support vector machine (SVM), k-nearest neighbor (KNN) and naïve Bayes. Our findings revealed that the ANN method has the best performance with a high level of accuracy. Demographic, academic, family/personal and internal assessment were found to be the most frequently used attributes in prediction the student performance.