 | ||
Validation Machine Learning Models To Predict Score On Graduate Tests Based On High School Test And Other Factors, Case Study: Colombia. | ||
| Published in: | Prospective and trends in technology and skills for sustainable social development. Leveraging emerging technologies to construct the future: Proceedings of the 19th LACCEI International Multi-Conference for Engineering, Education and Technology |  |
| Date of Conference: | July 19-23, 2021 | |
| Location of Conference: | Virtual | |
| Authors: | Maryori Sabalza Mejia (Universidad Tecnológica de Bolívar, CO) Carolina Campillo Jimenez (Universidad Nacional Abierta y a Distancia, CO) Juan Carlos Martinez Santos (Universidad Tecnológica de Bolívar, CO) | |
| Full Paper: | #343 | |
Abstract:In Colombia, the state usually administers tests to evaluate the knowledge learned during high school and university. This test is the Saber 11 Test, and it applies at the end of high school. These tests are an indispensable requirement in admissions for the university. Students must take the Saber Pro Test as a grade requirement at the end of said studies, which assesses university quality. However, many of the students who performed well on the Saber 11 tests may fail or even never take the Saber Pro Test because many drop out before finishing their degree. Many conditions may affect, but the student of the socio-economic conditions is one of them. This research shows the validation of machine learning models to predict the Saber Pro Test results based on the results of the Saber 11 test according to a range. This range was a maximum period of five years, considering socio-economic variables that remained constant during this time. Two models were verified that comply with a 100% prediction with the real value, and by the stacking model, the prediction values are correct up to 80.41%. | ||