Journal of Artificial Intelligence1994-54502077-2173orgz10.3923/jai.2009.65.72FahimifardS.M.SalarpourM.SabouhiM.ShirzadyS.2200922It is well documented that many economic time series observations are nonlinear and nonlinear models estimated by various methods can fit a data base much better than linear models. Beside they can learn from examples, are fault tolerant in the sense that they are able to handle noisy and incomplete data, are able to deal with non-linear problems and once trained can perform prediction and generalization at high speed. Therefore, in this study, the utilization of Adaptive Neuro Fuzzy Inference System (ANFIS) as a nonlinear model and Auto-Regressive Integrated Moving Average (ARIMA) model as a linear model are compared to agricultural economic variables time series forecasting. As a case study the three horizons (1, 2 and 4 week ahead) of Iran’s poultry retail price are forecasted using the two mentioned models. The results of using the three forecast evaluation criteria state that, ANFIS model outperforms ARIMA model in all three horizons. And consequently the effective role of ANFIS model to improve the Iran’s poultry retail price forecasting accuracy can’t be denied.]]>Box, G.P.F. and G.M. Jenkins,1978Chen, X., J. Racine and R.N. Swanson,2001Gencay, R.,1999Hann, T.H. and E. Steurer,1996Ho, S.L. and M. Xie,1998Ince, H. and T.B. Trafalis,2006Kalogirou, S.A.,2003Kamwa, I., R. Grondin, V.K. Sood, C. Gagnon, V.T. Nguyen and J. Mereb,1996Lapedes, A. and R. Farber,1987Makridakis, S.,1982Makridakis, S., S.C. Wheelwright and R.J. Hyndman,1998Racine, J.S.,2001Sharda, R. and R. Patil,1990Sugeno, M. and G.T. Kang,1988Wu, B.,1995Zhang, G. and M.Y. Hu,1998Haoffi, Z., X. Guoping, Y. Fagting and Y. Han,2007