Artificial Neural Network Model to Predict Academic Results in Mathematics II

Abstract

Objective. This article shows the design and training of an artificial neural network (ANN) to predict academic results of Civil Engineering students of the Fabiola Salazar Leguía National Intercultural University, from Bagua-Peru, in the subject of Mathematics II. Method. The CRISP-DM methodology was used, surveys were conducted to collect the data, and the RNA model was implemented in the Matlab software using the nnstart command and two learning algorithms: Scaled Conjugate Gradient (SCG) and Levenberg-Marquardt (LM). The performance of the model was evaluated through the mean square error and the correlation coefficient. Conclusions. The LM algorithm achieved better prediction effectiveness.