Model based on Machine Learning to alert early on the academic performance that a Systems Engineering student would have upon graduation

Zelada-Valdivieso H. · Proceedings of the 2021 IEEE 1st International Conference on Advanced Learning Technologies on Education and Research, ICALTER 2021 (2021) · DOI: 10.1109/icalter54105.2021.9675117

Resumen

Academic performance is very important evidence that shows the quality of educational programs, because many interest groups expect universities to adequately prepare their students. However, there are different studies that show that recent graduates do not have the skills required by the labor market. In this sense, a review of the literature was made regarding technological solutions related to alerts, prediction or early measurement of the graduate's academic performance, finding that the vast majority of research focused on predicting academic performance in classroom or classroom courses. virtual or in the prediction of student dropout, but the issue of early prediction of the academic performance that a student should upon graduation has not been studied much. Due to this, a computational model based on Machine Learning was developed that allows early warning of the academic performance that a Systems Engineering student would have upon graduation and in this way improvement actions can be proposed before the students leave the university. For this, the CRISP-DM methodology were used, arriving to determine that the model trained with Ensemble (Subspace discriminant) obtained greater precision: 81.4%.

Instituciones: Universidad Católica Santo Toribio de Mogrovejo