forecasting the export of Iran's date Using econometric methods and artificial intelligence

Subject Areas : Agricultural Economics Research

A. اکبری ¹ , M. Sh ² , H. مهرابی بشرآبادی ³

1 - استاد اقتصاد کشاورزی دانشگاه سیستان و بلوچستان
2 - کارشناس ارشد اقتصاد کشاورزی دانشگاه سیستان و بلوچستان
3 - دانشیار اقتصاد کشاورزی، دانشگاه شهید باهنر کرمان.

Received: 2014-04-13 Accepted : 2014-04-13 Published : 2013-11-22

Keywords: Iran, Artificial Neural Networks, Forecasting, ARIMA, GARCH, Date's export, Genetic algorithms,

Abstract :

This study attempts to forecast the export of Iran's date for the period of  2010-2016, using the econometric methods, ARMA, GARCH, and the methods of computational intelligence, artificial neural networks and genetic algorithms.The data from 1967-2009 were used. The data from 1967- 2005 were used for modeling and the four last years, were used to examine forecast ability. Results indicated that the neural network has the lowest Forecast error comparing other methods followed by Genetic Algorithm .According to the superiority of neural networks in forecasting the Iran's date export, forecasts made by this, shows the decreasing-increasing trend in the exports of Iran's date.