Iran’s Non-Oil Export Forecasting Using Adaptive-Neuro Fuzzy Inference System (ANFIS)

Document Type : Original Article

Authors

Abstract

This study introduces the Adaptive Neuro-Fuzzy Inference System (ANFIS) and compares its application in forecasting the three perspectives (1, 2 and 4 year(s)) ahead of Iran’s non-oil export -which the agricultural products made its main share with ARIMA as the most common econometric linear forecasting method during 1959-2008. Results of models performance evaluation showed that the forecasted test data related to ANFIS designed architects have more correspondence with the real data in comparison with the ARIMA forecasted out of sample data. Therefore the non-linear ANFIS model is more efficient than the linear ARIMA model for all considered perspectives of Iran’s non-oil export.

JEL Classification: C22, C45, E47, F31

Keywords:
ANFIS, ARIMA, Forecasting, Non-Oil Export