Applying a hybrid model of neural network and decision tree classifier for predicting university admission

Abstract

Predicting university admission is a complex decision making process that is more than merely relying on test scores. It is known by researchers that students' backgrounds and other factors correlate to the performance of their tertiary education. This paper proposes a hybrid model of neural network and decision tree classifier that predicts the likelihood of which university a student may enter, by analysing his academic merits, background and the university admission criteria from that of historical records. Our prototype system was tested with live data from sources of Macau secondary school students. In addition to the high prediction accuracy rate, flexibility is an advantage as the system can predict suitable universities that match the students' profiles and the suitable channels through which the students are advised to enter. Our model can be generalized with other attributes and perform faster when compared to using a neural network alone.