Forecasting of Agricultural Crops Import in Iran Artificial Neural Network and Econometric Models Application

Hosseini, S.S.; Pakravan, M.R.; Gilanpour, O.

doi:10.22067/jead2.v1390i3.10845

Forecasting of Agricultural Crops Import in Iran Artificial Neural Network and Econometric Models Application

Document Type : Research Article

Authors

S.S. Hosseini ¹

M.R. Pakravan ¹

O. Gilanpour ²

¹ agricultural Economics Department, Tehran University, Iran

² Foreign Trade & Marketing Research Department; Agricultural Planning, Economics and Rural Development Research Institute

https://doi.org/10.22067/jead2.v1390i3.10845

Abstract

Abstract
In the present study, agricultural sector import was forecasted by using the econometric and the ANN methods. Import data from 1971 to 2004 and 2004-2009 was used for forecasting, network training and testing forecast accuracy, respectively. The results shown that Feed-forward neural network has much less error and better performance than the ARIMA and the VAR methods. On the basis of the obtained results, import would be increased, in 2010-2013 while the increase in import will not be significant in 2014, it will increase again in 2015. Therefore fundamental efforts should be done to increase production potential and to achieve self-sufficiency, by essential triggering policies.

Keywords

Iran

Forecasting

Neural Network

Agricultural Crops

Import