Comparison of ANFIS, ANN and ARIMA Models Performance to Forecasting Case study: Weekly Egg Price

  As the precise price forecasting in agricultural economic productions has the main effect on policy making in this sector, using more effective models such as Artificial-Neural models can contribute the policy makers of this sector in future decision making. Therefore in this study the utilization of new ANFIS model for forecasting, was compared with ANN and ARIMA models. As an empirical application, the various forecasting performance of mentioned models for three horizons of weekly egg price were compared via common forecast performance measures. Also, the required data for the period 2002:3-2008:6 was obtained from the website of Iran State Livestock Affairs Logistics. Results showed that ANFIS and ANN models outperform ARIMA model, and ANFIS model outperforms ANN model for all three forecasts.     JEL Classification : C22, C8, L66