MODELING AND FORECASTING OF GROSS DOMESTIC PRODUCT PERCENTAGE SHARE OF EDUCATION SECTOR: A STATISTICAL STUDY IN PAKISTAN

Author(s) : MUHAMMAD WASIM AMIR, AYESHA BIBI, NOUREEN AKHTAR, ZEESHAN RAZA

Abstract:

Education is a very important medium for obtaining skills and knowledge. The world is continually
changing so it is significant to teach and bring up those peoples who can understand and solve modern social
problems. Forecasting techniques are necessary to find out the future trends that are meet the upcoming problems.
For the desired purpose, different time series models are applied and selected the best one for better-forecasted
values of the gross domestic product (GDP) percentage share on education. The best model is determined based
on the lowest value of the Akaike information criterion (AIC) and Bayesian information criterion. Moreover,
the quality and predictability power of the selected models are measured based on the minimum value of the mean error, root mean square error, mean percentage error, and mean percentage absolute error. The AIC criteria showed that the best time series model is the autoregressive integrated moving average (ARIMA) (2, 1, 1) to forecast the (GDP) percentage share on the education of Pakistan. Furthermore, the assumptions of the fitted model such as independence, no heteroscedasticity, and normality are also evaluated. The selected model showed that the (GDP) percentage share on education is decreasing slowly, which is not good for the education sector and development of the country.