Comparative Analysis of Machine Learning Techniques Student Final Grade Classification Accuracy

In education, assessment is often used as an indicator of student success during the teaching-learning process. Although schools provide facilities and teaching staff with a uniform portion, the gap in grades still often occurs between students in each educational institution. This research aims to implement and classify the performance of machine learning algorithms to predict student’s final grades and analyze related factors in secondary level students using Student Alcohol Consumption data from UC Irvine Repository. The data consist of 650 rows with 33 attributes and there are 5 types of grading systems. This research utilizes machine learning methods, specifically Naive Bayes, Neural Network MLP, Logistic Regression, Gradient Boosting, Gaussian Process, Linear Discriminant Analysis, SVM, XGBoost, Random Forest, and KNN. The results showed that the classification model built using KNN has the highest accuracy compared to other classification models with an accuracy of 82.31%. The contribution of this research is to improve the classification accuracy results from previous research using 10 classic machine learning algorithms. With these results, this research can be continued into an application that can classify student performance and predict final grades more accurately.