Mohammed Goli Yousefi; Timur Mohammadi; Navid Maarefzadeh
Volume 2, Issue 7 , July 2013, , Pages 147-170
Abstract
The purpose of this paper is to forecast demand for crude oil of Iran using Artificial Neural Networks and ARMAX models. The result indicates that Artificial Neural Networks provides an accurate and better picture compared with ARMAX. In order to show whether the variables used in this study are true ...
Read More
The purpose of this paper is to forecast demand for crude oil of Iran using Artificial Neural Networks and ARMAX models. The result indicates that Artificial Neural Networks provides an accurate and better picture compared with ARMAX. In order to show whether the variables used in this study are true determinants of Crude oil demand, we have also applied the same techniques with the same variables to forecast crude oil demand of five selected OPEC countries. The result confirms our earlier findings for Iran. Applying rank correlation coefficient for these findings, show high correlation coefficients between the result for Iran and other countries. Therefore we may say that the variables such as GDP, population, net exports and the number of vehicles are key variables for any forecasting relating to crude oil demands in similar countries.
