Identificación de factores en la reprobación y deserción mediante técnicas de minería de datos en el Área de la Energía de la Universidad Nacional de Loja.

Currently the Higher Education Institutions face a complex problem, this problem is the college dropout and emerges from the student starts school and then passes through the National System of Equalization and Admission (SNNA), and then studying for a career. This has been used a lot of efforts in order to have students work directly or enhance their academic commitment. For countering this problem a number of activities have been evidenced such as workshops, extra class, tasks, among others. Without achieving improve the situation, it means, these efforts have not been sufficient and the problem still persists in educational higher-level institutions Based on writing above, the main objective of this work is to identify the factors that influence on university dropout and fail. To achieve this process were applied to data mining techniques that will allow drawing conclusions and relevant information on dropout and failure rates. The selection of data mining techniques is justified by literature sources; they are essentially for this study. By other hand the methods used to achieve, the efficient development of this work has used the CRISP-DM methodology, which gives the possibility to bring an orderly and iterative work. The data used to carry out the identification of the factors was obtained from the databases of SISTEMA DE GESTIÓN ACADÉMICA (SGA) through its Web services also details University Welfare Area was collected, and then integrate this on a single source of data for later processing. Finally, after identifying factors dropout and failure, a predictive model of attrition in order to validate the results was obtained, which was evaluated with data from students currently sign on of Energy Industries and Natural Resources Non-Renewable carriers of Universidad Nacional de Loja.

• 1. autor: González Pineda, Anibal Israel (author)
• Format: bachelorThesis
• Język: spa
• Wydane: 2014
• Hasła przedmiotowe: "<REPROBACIÓN DESERCIÓN> <TÉCNICAS DE MINERÍA DE DATOS> <HERRAMIENTAS ÚTILES EN MINERÍA DE DATOS> <METODOLOGÍA CRISP-DM>"
• Dostęp online: http://dspace.unl.edu.ec/jspui/handle/123456789/14074