Modelo de retención universitaria: Un enfoque de Machine Learning.

University retention has become a globally recognized phenomenon, due to its complexity and multiple causes that must be addressed in the university environment; the decrease in its rates generates academic and management difficulties for Higher Education Institutions. It is considered important to analyze retention issues as a means of mitigating problems that affect the student and that allow the successful completion of a career. For this reason, a model is proposed to identify university student retention factors based on the application of Machine Learnign techniques. The data is obtained from an information survey process through a survey of 294 students from a public university in Ecuador, for the development of research the Knowledge Discovery in Database (KDD) methodology and supervised learning algorithms with neural networks are used. The results allow to design a conceptual model based on 7 factors that influence the retention of students in universities in universities using Linear Regression. For the prediction, Cluster and Neural Networks algorithms were used, resulting in an accuracy rate of the proposed models of 94.2% with the Multilayer Perceptrom model, which allows to determine that the research developed is based under the experimental procedure that checks the validity of the proposed conceptual model.

• Hlavní autor: Urgiles Urgiles, José Luis (author)
• Další autoři: Vásquez Mullo, Marcia Salome (author)
• Médium: bachelorThesis
• Jazyk: spa
• Vydáno: 2020
• Témata: RETENCIÓN DE ESTUDIANTES; MACHINE LEARNING; CLÚSTER; REDES NEURONALES; SISTEMAS
• On-line přístup: http://repositorio.utc.edu.ec/handle/27000/8608