Abstract

RICO PAEZ, Andrés  and  SANCHEZ GUZMAN, Daniel. Design of a model to automate the prediction of academic performance in students of IPN. RIDE. Rev. Iberoam. Investig. Desarro. Educ [online]. 2018, vol.8, n.16, pp.246-266. ISSN 2007-7467.  https://doi.org/10.23913/ride.v8i16.340.

Educational data mining allows extracting useful and understandable knowledge from academic data to solve problems about various teaching and learning processes. One of the most popular applications of educational data mining is the prediction of academic performance. The main objective of this work was to design and automate a predictive model of the academic performance of students of the National Polytechnic Institute (IPN).

For the construction of the model, the qualifications of five academic activities and the final grade of 94 students enrolled in an Engineering career belonging to the IPN were analyzed. This model was applied to 86 students to predict their academic performance. Subsequently, these predictions were compared with the actual results obtained by the students at the end of the course. Accuracy was obtained from the predictions of the course approval of up to 73% and only with five attributes corresponding to the qualifications of the initial academic activities. In addition, a platform was built that facilitates the construction and use of the model to automatically predict the academic performance of new students. Also, the main academic activities that influenced academic performance were identified through the value of the probabilities of the model. In particular, the results showed that activities 3, 4 and 5 were those that most significantly influenced the prediction of approval of the students who participated in this study. The development of this type of models allows educational institutions to predict the academic performance of their students and identify the main factors that influence it.

Keywords : Naïve Bayes algorithm; data mining; predictive model; probabilities; academic performance.

        · abstract in Spanish | Portuguese     · text in Spanish     · Spanish ( pdf )