Multi-layer Adaptive Fuzzy Inference System for Predicting Student Performance in Online Higher Education

Research on student performance prediction has evolved from the early application of statistical techniques to later use of computational techniques. Results in this field are varied, thus, we have to take advantage of previous research results. This study proposes a Multi-layer Adaptive Neuro-Fuzzy Inference System (MANFIS) for student performance prediction in online Higher Education settings. The MANFIS was trained and tested using a dataset integrated by the scores obtained by students in four online Higher Education courses. The MANFIS prediction accuracy was compared against the accuracies of Multilayer neural network, Radial Basis Function Neural Network, and General Regression Neural Network. The accuracy of the MANFIS prediction statistically outperformed at least one neural network (out of three possible) in each dataset. The Results indicate that MANFIS is an alternative model to predict student performance in online Higher Education settings.