Diseño de un modelo matemático mediante técnicas de análisis multivariado para estimar la deserción estudiantil en el Instituto Tecnológico Superior Cotacachi.

The following degree work, called “Design of a mathematical model through multivariate analysis techniques to estimate student desertion in the Technological Superior Institute Cotacachi”, has as objective to develop a predictive model that allows to estimate the rate of student desertion in the institution. The research analyses socioeconomic, academic and demographic facts that influence in desertion, applying multivariate analysis techniques for identification and quantification. The methodology used includes data collection of 1.385 records of enrolled students between 2022 and 2024, followed by a descriptive analysis and the application of a model of multivariate logistic regression. This model was validated through performance metrics such as precision recall and F1-score, obtaining an accuracy of 79%. The results highlight that variables like age, repetition of subjects and loss of free access influence significantly in desertion, while facts like sex, marital status were not drastically relevant. In conclusion, the predictive model proposed responds to the objective of identifying with precision to the students at risk of desertion, allowing the implementation of early intervention strategies.

• المؤلف الرئيسي: Estrada Arana, Juan Pablo (author)
• التنسيق: masterThesis
• اللغة: spa
• منشور في: 2025
• الموضوعات: deserción estudiantil; análisis multivariado; regresión logística; modelo predictivo; educación
• الوصول للمادة أونلاين: http://dspace.unach.edu.ec/handle/51000/14558