Classification model for student attrition in a Peru public university

Torres H.V.; Rodriguez W.M.; Morales J.Á.; Mejía J.C.; Murillo C.M. · Salud, Ciencia y Tecnologia - Serie de Conferencias (2023) · DOI: 10.56294/sctconf2023175

Resumen

Information and communication technologies are playing a very important role in the different fields of knowledge; there is an increasing ability to identify patterns in the data of an organization; in this context, the study aimed to develop a classification model for student dropout by applying the autoML method of the H2Oai framework, taking into account socioeconomic and academic characteristics of students in order for authorities to make decisions in a timely manner. The methodology was technological, propositional level, incremental innovation, temporal scope, synchronous; the data collection was prospective, a 20-item questionnaire was applied to 237 graduate students of master’s degree programs in education. The research resulted in a supervised automatic learning model, Gradient Boosting Machine (GBM), to classify student desertion, identifying the main associated factors that influence desertion, obtaining a Gini coefficient of 92,20 %, AUC of 96,10 % and a LogLoss of 24,24 %. © Autor(es); 2023.

Instituciones: Universidad Nacional José Faustino Sánchez Carrión