Grade Prediction Based on Student Performance using Gradient Boosting and Random Forest

Predicting the academic performance of students is fundamental to educational. Schools can identify learning deficits and developments for timely interventions. This study is used to predict students’ grades using Gradient Boosting and Random Forest machine learning methodologies. This study used a dataset that focused on students’ performance and other relevant factors, the dataset this study used includes such variables as Grade Point Average (GPA), absences, parental support, parental education, tutoring, weekly study time, extracurricular, and sports. According to the evaluation results, the Gradient Boosting has a precision of 0.929 and it is a good result. Besides that, in predicting students’ final grade results, the Gradient Boosting model performed better than Random Forest model. With these predictive models, the educational institution will be able to predict academic performance of students with higher accuracy and provide academic for help and support for student learning. According to the results the Gradient boosting model can be a crucial tool for educational institutions to increase the quality of learning education and help students perform better.