APPLICATION OF MACHINE LEARNING TECHNIQUES IN AN ACADEMIC CONTEXT WITH A FOCUS ON IDENTIFYING DROPOUT AND DROPOUT STUDENTS

School dropout is one of the main problems that cause losses to Higher Education Institutions and the use of prediction models con subsidize decisions to minimize losses. In this context, this work evaluates whether it is possible to employ machine learning algorithms to generate modeling the evasion pattern, based on academic record data. This hypothesis was validated through a case study, using academic data from the State University of Tocantins. The results achieved by the experiments indicated that the methodology adopted in this work was able to classify students in situations of evasion and non-evasion with a high degree of confidence.