Monitoring School Failure from a Student Performance Prediction Model
Research indicates that, in recent decades, there has been a consistent expansion in the number of students entering higher education in Brazil. In recent years, this growth has been sustained, especially by the numbers of the distance modality. However, several problems associated with student performance and retention remain. Artificial Intelligence in Education and, more specifically, machine learning techniques have been used to deal with dropout problems and students' academic performance. A systematic review of the literature identified that, although there are works that present high accuracy values in predicting student performance, there still needs to be more in relating and showing the benefits of this prediction for students and teachers. In this sense, this work aims to create a prediction model with a high power of discrimination between the different performances of the students. The expectation is that the model can support the detection of school failure during a discipline so that the teacher and the student can carry out interventions to reverse the predicted scenario.