Diseño de políticas de apoyo escolar dirigidas a estudiantes de primaria con alto riesgo de repetición: Un enfoque basado en modelos de aprendizaje automático

Abstract

In recent years the Uruguayan education system has positioned itself among the worst in the region in terms of completion of secondary education and repetition in primary education. Despite the fact that there is enough empirical evidence reflected in a long discussion in articles and specialized journals that these two phenomena are related, most of the actions taken to prevent disengagement from formal education are aimed at students from the middle education level. By the use of statistical models and machine learning this work aims to identify who are these primary school students who have a higher repetition risk, in order to be able to take the most appropriate prescriptive actions. These actions would help prevent school dropout and have a positive impact, so that students at risk could successfully complete the primary education cycle and also have greater possibilities of finishing the secondary and higher education cycle.