Application of the ARIMA Models for Predicting Students’ Admissions in the University of Lagos
No Thumbnail Available
Journal of Scientific Research and Development
The objective was to assess the performance of the AutoRegressive Integrated Moving Average (ARIMA) models when occasional level shifts occur in the time series under study. The secondary data on the University of Lagos’ undergraduates’ admissions (1962–2016) were collected and analysed. It is predicted that universities in Nigeria and elsewhere could forecast their enrolment figures and student population growth rate based on the ARIMA models. The Box–Jenkins (B–J) approach provided the theoretical framework for the statistical analysis. The study used the Kalman Filter (KF) algorithm to develop a method using an ARIMA model to overcome and resolve the three main problems of the B–J methodology. The KF estimated the states for dynamic systems in state-variable formulations. Forecasting university admissions is necessary if student population must match the infrastructural provisions on campuses. The best ARIMA models have been selected by using criteria such as Akaike’s Information Criterion (AIC), Schwarz’s Bayesian Criterion (SBC), Absolute Mean Error (AME), Root Mean Square Error (RMSE) and Mean Absolute Percent Error (MAPE). To select the best ARIMA model, the data was split into two periods: estimation period and validation period. The results clearly showed a continual increase in the demand for university education in the University of Lagos and, by extension, other universities in Nigeria.
AutoCorrelation Functions (ACF) , Forecast , AutoRegressive Integrated Moving Average , Algorithmn , Research Subject Categories::NATURAL SCIENCES