Predicting Student Dropouts in Higher Education Using Supervised Classification Algorithms

The aim of this paper is to predict, on a purely algorithmic basis, students who are at risk of dropping out of university. Data used in this study originated from the University of Bari Aldo Moro, during 2013–16, and were provided by the Osservatorio Studenti-Didattica of Miur-Cineca. Data analysis is based solely on the information set available, for each student, inside the university information system. Predictions of individual dropouts have been carried out by means of suitable Machine Learning techniques, known as supervised classification algorithms.