An Adaptive Neuro-Fuzzy Inference System (ANFIS) Model for Prediction of University Program Recommendation

Sarerusaenye Ismail, Shahrinaz Ismail and Nur Haliza Abdul Wahab

It is observed that a few expert systems implemented at the universities in Malaysia are only to advise students for course selection, but none are implemented to recommend university programs to potential and new students. Recent research has proven the existing issues among students who have mistakenly chose and undertaken unsuitable programs that led to study failures, such as drop out cases poor academic achievements and performances. There are many factors that influence these failures, which are categorised into internal and external factors. These factors are poor guideline during consultation session in the initial enrolment, lack of information on choosing the university program, influences from family, friends, agents and peers, and lastly the university system itself. Hence, this study develops the ANFIS system to provide university program recommendation, in which by using the neuro-fuzzy logic technique, the university is able to learn the complexity of university program. MatLab has been selected for developing the fuzzy logic and neural network has been chosen to do data training and evaluation. The results obtained from the prediction of fuzzy logic and system testing indicate optimum and accurate target outputs.

Volume 11 | 10-Special Issue

Pages: 1100-1107

DOI: 10.5373/JARDCS/V11SP10/20192911