Multiclass Prediction Model for Student Grade Prediction Using Machine Learning

Abstract

Student grade is one of the key performance indicators that can help educators to monitor student’s academic performance. Predicting this student’s grade is a tedious task for the teachers. This paper presents a comprehensive analysis of machine learning techniques to predict the student grade. The accuracy performance of four machine learning techniques namely Decision Tree, Naïve Bayes (NB), K-Nearest Neighbor (KNN) and Random Forest (RF) is compared using student's course grade dataset. Multiclass prediction model is proposed to reduce the overfitting and misclassification results caused by imbalanced multi-classification based on oversampling Synthetic Minority Oversampling Technique (SMOTE) with forward features selection methods. The obtained results show that the proposed model integrates with RFgive significant improvement in accuracy with 98%. This model indicates the comparable and promising results for imbalanced multi-class dataset for student grade prediction. Keywords – kNN Imputer, SMOTE, Forward Feature Selection, k Nearest Neighbor, Naïve Bayes, Decision Tree, Random Forest.