Comparing regression using artificial neural nets and intelligent hybrid method to achieve the higher learning preferences of students

Abstract

This work compares an artificial neural net and computational intelligence hybrid method to describe the learning preferences of students from survey data. The final purpose is to give learning objects to students according to their learning style. We used a database form survey with answers from 1042 computational engineers students from two public Universities in Tijuana, Mexico. We also used the Fuzzy Inference System (FIS); the FIS is configured from survey data using the ANFIS method to discover the set up and fuzzy if-then rules of the system. The FIS describe learning objects preferences for learning styles; then we compared the results from ANN versus ANFIS, in order to retrieve the highest results to build learning objects.