Análise de Evasão no Ensino Superior na Modalidade EAD Utilizando Estatística e Machine Learning

Authors

Marlos Vinicius Piccinelli, Universidade Tecnológica Federal do ParanáFollow
Marcelo Gonzaga, Universidade Tecnológica Federal do ParanáFollow
Luiz Celso Gomez Júnior, Universidade Tecnológica Federal do ParanáFollow
Nádia P. Kozievitch, Federal University of Technology - ParanáFollow
Rita Cristina Galarraga Berardi, Universidade Tecnológica Federal do ParanáFollow

Abstract

Student dropout is a significant challenge for educational institutions, affecting approximately 40% in Higher Education. This study analyzes data from students enrolled in Distance Education (EAD) courses at a Higher Education Institution (HEI). It employs statistical techniques and Machine Learning (ML) to categorize students and identify variables that influence dropout. Through data visualization, Logistic Regression, and Clustering (K-Means), the study identifies key factors affecting dropout and develops a model for grouping students based on their academic profiles. The authors also create a descriptive text for each group, facilitating the identification of students with a high-risk profile. However, limitations of the study include incomplete data input by the HEI, with important fields left unfilled.

Recommended Citation

Piccinelli, Marlos Vinicius; Gonzaga, Marcelo; Gomez Júnior, Luiz Celso; Kozievitch, Nádia P.; and Galarraga Berardi, Rita Cristina, "Análise de Evasão no Ensino Superior na Modalidade EAD Utilizando Estatística e Machine Learning" (2024). ISLA 2024 Proceedings. 1.
https://aisel.aisnet.org/isla2024/1