Modelo de aprendizaje profundo para la predicción de deserción de estudiantes de la carrera de informática

Abstract

Abstract: This study addresses student dropout in distance university education, especially in the Computer Science degree at UTPL. The main causes are examined and the application of artificial intelligence and deep learning in analysis and dropout prediction. The work proposes a model based on deep learning to identify attrition patterns and evaluate model accuracy. It follows the KDD (Knowledge Discovery in Databases) methodology to develop the predictive model and emphasis is placed on the importance of addressing the dropout problem given its negative consequences. During development, the agile Scrum methodology was used for its benefits of iterative and collaborative work. Specific data was collected from the students and a dashboard was designed for an effective visualization of the results of the model, facilitating the interpretation of complex data.