Modelo de análisis del rendimiento académico de la Unidad Educativa Personas Con Escolaridad Inconclusa. (P.C.E.I.) “Monseñor Leonidas Proaño” del cantón Latacunga, a través de minería de datos.
The main objective of this work is to contribute to the process of predicting the academic performance of the students of Unidad Educativa de Personas Con Escolaridad Inconclusa (PCEI) Monseñor Leonidas Proaño from Latacunga city by means of the integral study of techniques and tool of analysis of m...
שמור ב:
| מחבר ראשי: | |
|---|---|
| פורמט: | masterThesis |
| שפה: | spa |
| יצא לאור: |
2020
|
| נושאים: | |
| גישה מקוונת: | http://repositorio.utc.edu.ec/handle/27000/7142 |
| תגים: |
הוספת תג
אין תגיות, היה/י הראשונ/ה לתייג את הרשומה!
|
| סיכום: | The main objective of this work is to contribute to the process of predicting the academic performance of the students of Unidad Educativa de Personas Con Escolaridad Inconclusa (PCEI) Monseñor Leonidas Proaño from Latacunga city by means of the integral study of techniques and tool of analysis of mining of data from the factors of influence such as the social, economic and academic, determining indicators that detect elements which will serve the teachers, authorities and educational mentors to improve the academic performance of the student in the process of education. One of the stages of this research was to design a theoretical model of student retention through the Statistical Package for Social Sciences (SPSS) software applying linear regression, ordinary least squares that allowed to create the theoretical model of academic performance. This model followed an experimental process with four classification algorithms through machine learning techniques such as J48, Random Forest, Naive Bayes and OneR, this process was used to predict the precision rate of the proposed model. The implementation of these techniques allowed us to determine that the Naive Bayes algorithm presents an accuracy rate of 88.85%, which indicates that the model presented is adequate in terms of reliability, the layer levels obtained through the experimental process with a result of 0.86 indicate that these models are adequate to predict students retention. |
|---|