Comparative Analysis of the Prediction of the Academic Performance of Entering University Students Using Decision Tree and Random Forest

Aguilar-Ruiz J.; Taya-Acosta E.; Taya-Osorio E. · Lecture Notes in Networks and Systems (2023) · DOI: 10.1007/978-3-031-33258-6_43

Resumen

In this work, we discuss the most determinant variables in predicting the first-grade university courses (period 2021-I) of students who course a degree in the Faculty of Engineering in the Jorge Basadre Grohmann National University in times of pandemic. We used Machine Learning algorithms to determine the relations between high school grades, students’ personal information and preferences, and their final grade in the first semester of 2021. Our research found that the main variables are the modality of university entrance, students’ self-perception in the subject of leadership, the frequency of watching TV, YouTube, and Netflix, and the literary genres prefer. To confirm our results, we got an accuracy of 93.28% with DT and 93.33% with RF.

Instituciones: Universidad Nacional Jorge Basadre Grohmann