1 option
TeÌcnicas de mineriÌa de datos para determinar la desercioÌn escolar
- Format:
- Book
- Author/Creator:
- Apaza-Tarqui, Alejandro
- Language:
- Spanish
- Physical Description:
- 1 electronic resource (125 p.)
- Place of Publication:
- Puno Instituto Universitario de InnovacioÌn Ciencia y TecnologiÌa Inudi PeruÌ 2022
- Language Note:
- Spanish
- Summary:
- The objective of this research was to determine the data mining techniques and the associated factors that allow the segmentation of students at risk of dropping out at the Instituto Superior TecnoloÌgico Privado ISTEPSA, in Andahuaylas (Peru). For this purpose, Automatic Learning and Data Mining techniques implemented in WEKA software were applied: The CfsSubsetEval evaluation method and the BestFirst search method were applied to select the most significant factors, to establish the patterns the association algorithm A was used. priori and to segment, the Expected Value Maximization algorithm "Expectation Maximissation" (EM) and Kohonen's self-organizing maps (Self Organizing Maps, SOM) were used. The following results were obtained: 06 significant factors: Motivation of sessions, Laboratories and Classrooms of the Institution, Acceptance of the professional career, Repeated Courses in the school and Academic Semester; For dropout patterns, 100% of students who dropout rate motivation, classrooms, and laboratories as deficient; In addition, 96% consider the professional career they are studying to be deficient and 90% of those who withdraw are from the fourth semester; In the segmentation, 3 groups have been constructed with the EM algorithm and 4 groups for the SOM algorithm, where it is observed that the academic factors are decisive for the dropout of students.
- La presente investigacioÌn tuvo por objetivo determinar las teÌcnicas de mineriÌa de datos y los factores asociados que permitan segmentar los alumnos con riesgo de desercioÌn en el Instituto Superior TecnoloÌgico Privado ISTEPSA, en Andahuaylas (PeruÌ). Para este fin se aplicaron teÌcnicas de Aprendizaje AutomaÌtico y MineriÌa de Datos implementadas en software WEKA: Se aplicoÌ el meÌtodo de evaluacioÌn CfsSubsetEval y el meÌtodo de buÌsqueda BestFirst para seleccionar los factores de mayor significancia, para establecer los patrones se usoÌ el algoritmo de asociacioÌn A priori y para segmentar, se usoÌ el algoritmo de MaximizacioÌn del Valor Esperado "Expectation Maximissation" (EM) y mapas auto organizados de Kohonen(Self Organizing Maps, SOM). Se obtuvo los siguientes resultados: 06 factores significativos: MotivacioÌn de sesiones, Laboratorios y Aulas de la InstitucioÌn, AceptacioÌn de la carrera profesional, Cursos Repetidos en el colegio y Semestre AcadeÌmico; para los patrones de desercioÌn el 100% de los estudiantes que se retiran califican como deficiente la motivacioÌn, aulas y laboratorios; ademaÌs el 96% consideran deficiente a la carrera profesional que estudian y 90% de los que se retiran son de cuarto semestre; En la segmentacioÌn se ha construido 3 grupos con el algoritmo EM y 4 grupos para el algoritmo SOM, donde se observa que los factores acadeÌmicos son determinantes para la desercioÌn de alumnos.
The Penn Libraries is committed to describing library materials using current, accurate, and responsible language. If you discover outdated or inaccurate language, please fill out this feedback form to report it and suggest alternative language.