16th Brazilian Meeting of Bayesian Statistics and VI Latin American Conference on Statistical Computing Proceedings

Due to the necessary measures to prevent the spread of COVID-19, many students were forced to take distance classes. This research work focuses on characterizing students of an educational program of the Autonomous University of the state of Hidalgo who are vulnerable in their academic performance in the face of forced change in the modality of studies. The characterization of these vulnerable students is carried out using two data mining techniques that are developed with data corresponding to the sociodemographic, economic, technical, health, and academic conditions of each student in the sample set. The first data mining technique used shows the existence of two groups (clusters) in which differences can be observed, which makes it possible to differentiate that one group is in better conditions than another. The second technique allows making a list of the conditions that favor both positively and negatively for a student to see their academic performance affected.