Logistic model of desertion using regression and decision tree techniques for resource allocation efficiency : The case of a Chilean private university

Université Pontificale Catholique du Chili – Chili
Université Bernardo O’Higgins – Chili
Centre de Recherche Institutionnellebastian.gutierrez@ubo.cl

Roberto CORTÉS
Université de Salamanque – Espagne
Université Bernardo O’Higgins – Chili
Centre de Recherche Institutionnelle roberto.cortes@usal.es

Université Bernardo O’Higgins – ChiliCentre de Recherche Institutionnelle

Abstract : When education has become a fundamental right that must be maintained and consolidated, a problem arises that has given rise to numerous academic discussions and that seeks to consider holistically and critically the phenomenon of the growing massification of training offers, manifested in the exponential increase in access to postgraduate studies. Some approaches postulate that, not only is it necessary to implement policies to increase access coverage, but it is also necessary to address the quality of education, considering desertion and retention. In this context, this research aims to establish an analytical model that allows the development of retention and the prevention of the causes of desertion. For this purpose, we will use the methodology of supervised learning to determine the variables of analysis through regression techniques and a decision tree, creating a logistic model of desertion, capable of improving the efficiency in the destination of resources. Finally, this work will aim to provide a better understanding of the phenomena associated with the desertion and retention of students, to help in the management and decision-making process by institutions of higher education.

Keywords : higher education system, retention, desertion, logistic model.

JEL classification : C02, I23.

DOI : https://doi.org/10.18559/RIELF.2023.1.6