PREDICTING STUDENTS«¤?? ENROLLMENT USING GENERALIZED FEED-FORWARD NEURAL NETWORK

Abstract

An important obligation of educational planning is the projection of students«¤?? enrollment which forms the basis for many of the investment decisions. Enrollment projection provides information for decision making and budget planning hence, it is important to the development of higher education. As many factors have impacts on the enrollment number, and for the above reasons, students«¤?? population and enrollment number should be considered as a chaotic system. In this research, a Generalized Feed-Forward Neural Network (GFFNN) for students«¤?? enrollment prediction was proposed. The architecture of the proposed model was in-line with eight steps involved in developing a neural network model for predicting a chaotic system. The data used was obtained from Academic Planning and Quality Control Unit of Tai Solarin University of Education, Ogun State Nigeria. The results from the study showed that the mean absolute percent error of GFFNN has an average of 0.0101% unlike linear regression and autoregression models that were compared with it, with an average of 0.0570% and 0.0725% respectively. The proposed methodology is expected to assist the school management to adequately plan for the future needs of the students in the provision of facilities.ª¤?