Beschreibung
This research attempts to make comparison of the forecasting performance of regression models and time series models. The primary purpose behind this research was to find out which of the two models - regression model and univariate non seasonal ARIMA model - is more accurate and appropriate for forecasting purposes in the real world, keeping in view the cost of the model building. The research utilized data on Pakistans exports. The performance of regression models was compared with that of ARIMA models by using the statistics, TIC, RMSPE, MAE, MPE and MAPE. The comparison indicates that the ARIMA models perform much better than the regression models. It was also observed that the plots of actual values of the variables and those of the predicted values based on ARIMA model were closer than those of regression model. This result further supports that the ARIMA models perform better than the regression models. On the basis of the findings and discussion, it is strongly recommended that findings of the this research work may be implemented in policy making, especially in planning and development in developing countries like Pakistan because it is cost effective and more accurate.
Autorenporträt
Prof. Dr. Muhammad Afzal acquired Ph.D. degree in Economics from University of the Punjab, Lahore-Pakistan. He has presented papers in various peer reviewed national and international conferences. His area of interest is Applied Econometrics. Currently, he is serving as Professor of Economics at Lahore College for Woman University, Lahore.
Herstellerkennzeichnung:
BoD - Books on Demand
In de Tarpen 42
22848 Norderstedt
DE
E-Mail: info@bod.de
