Prediction of University Student Satisfaction in the Educational Experience Using Decision Trees and Logistic Regression: A Comparative Analysis

López-Cueva M.A.; Zegarra-Ugarte S.J.; Torres-Pacheco E.H.; Arpasi-Chura R.F.; Enríquez-Salas P.; López-Paz P.M. · Pakistan Journal of Life and Social Sciences (2024) · DOI: 10.57239/pjlss-2024-22.2.00591

Resumen

The objective of this study is to predict student satisfaction in higher education by applying decision tree and logistic regression models. A digital questionnaire was distributed to students at the National University of Altiplano, Peru, resulting in 1,020 valid responses. The dataset included 20 predictor variables, such as teaching quality, access to resources, and academic workload, alongside one dependent variable: student satisfaction. Decision trees and logistic regression models were trained with 80% of the data and tested with 20%. The results showed that the decision tree model achieved an accuracy of 95.6% based on the confusion matrix, while logistic regression reached 92.6%. Both models effectively identified key factors influencing student satisfaction, with the decision tree showing slightly better predictive performance. The study concludes that decision trees provide a more precise model for predicting student satisfaction in comparison to logistic regression. These findings highlight the usefulness of machine learning techniques in educational management for identifying areas needing improvement, ultimately enhancing student experiences and academic outcomes. This study provides a reliable model for university administrators to optimize resource allocation and improve educational practices. The use of machine learning models can lead to data-driven decisions that positively impact student satisfaction and institutional performance. © (2023), (Elite Scientific Publications). All Rights Reserved.

Instituciones: Universidad Nacional del Altiplano