Modelo para la predicción de la deserción de estudiantes de pregrado, basado en técnicas de minería de datos

The main objective of this research project is to create a model for the prediction of undergraduate student desertion at the Universidad de la Costa - CUC, based on the analysis of different socioeconomic and academic factors. The study required the execution of a series of phases: characterization, experimentation, development and evaluation. During the characterization phase, a dataset was constructed, based on the compilation of demographic, cultural, social, family, educational, socioeconomic status and psychological profile data of each student, for the periods between 2013-1 and 2018-2. Such information was collected from the registration forms that students fill out when they enter the institution, a total of 1,606 unique student records were collected. During the experimental phase, different machine learning techniques were evaluated for the categories: Bayesian networks, support vector machines, and decision trees. The algorithm with which the best hit rate was obtained was Random forest (from the decision tree category), with an accuracy of 84.8%. In the development phase, the model was integrated into an application that allows us to predict whether a student or a group of students will drop out or not. Finally, in the evaluation phase, the application was subjected to different types of tests to evaluate both the functionality of the graphic interface with the final user and the success rate in terms of desertion prediction, the results have coincided with the precision obtained in the experimental phase.

• Autores: Camargo García, Aníbal José
• Fecha de publicación: 2020
• Institución: Corporación Universidad de la Costa
• Repositorio: REDICUC - Repositorio CUC
• Idioma: spa
• OAI Identifier: oai:repositorio.cuc.edu.co:11323/7077
• Acceso en línea: https://hdl.handle.net/11323/7077; https://repositorio.cuc.edu.co/
• Palabra clave: Higher education; Dropout; Data mining; Decision tree; Classification; Prediction; Educación superior; Deserción; Minería de datos; Árboles de decisión; Clasificación; Predicción
• Rights: openAccess
• License: Attribution-NonCommercial-ShareAlike 4.0 International