investigating the export function of border markets in iran, using the ann approach

investigating the export function of border markets in iran, using the ann approach ali falahati faculty of economics razi university of kerman minoo nazifi naeini sahar abbaspour abstract nowadays, trade has been introduced as an engine of growth and development in developing countries so states are required to participate in international trade more seriously. one way for expanding participation in international trade is regional transactions. border trade is one of the main indicators in communicating between neighboring countries and the livelihood of the frontier. establishment of border markets makes the two countries expanding trade and avoid conflict and reduce political tensions with neighboring countries .so evaluating the economic performance of these malls is a good way for evaluating their situation. in this paper the export of border malls is evaluated with artificial neural networks we examine data from 1986 to 2010 in iran using artificial neural method.. in this study there are 5 independent variables that is defined as the entrance layer for neural networks. the results show that ann method has a small error less than 0.004 and it shows that ann is a good method for forecasting the ann method. in this method the more important variable is non petroleum exports.