Relative performance of statistical and fuzzy regression models in estimation of gasoline demand

The most challenges for gasoline are uncertainties resulting from structural failure of economy، changing of policies and lack of accurate data. The objective of this research is to specify the variables explaining the gasoline demand in Iran and to estimate gasoline demand by using the statistical and fuzzy regression models over the period 1981-2007. Finally، the models were compared using standard criteria. The results indicate that the both of methods have enough accuracy for estimating and forecasting of gasoline demand. In addition، there is a significant relationship between gasoline price، car per capita and trend as explanatory variables of gasoline demand per capita.