Modeling and Forecasting Method in seafood Export ARIMA and ArtificialNeural Networks

The main objective of this paper is to modeling and forecast exportsing of seafood in Iran. For this purpose, the method of collective self-explanatory time series moving average (ARIMA) and artificial neural networks are used. To perform the study, monthly data for the period of 1995:05 to 2008:02 estimated from model training period 2009:03 to 2011:02 data to verify the predictive power of the model. In this study, several criteria including Absolute Error (MAE), Mean Square Error (MSE), Average Percentage Error (MAPE), Root Mean Square Error (RMSE) and Root Mean Square Normalized Error (NRMSE) were used. The results show better performance of the neural network predicted non-linear statistical model ARIMA models and neural network structures studied neural networks Radial Basis Function (RBF) has the best performance in terms of error functions. Finally, for the two years 2012 and 2013 the amount of Iranian seafood export is predicted.