hybrid modeling approach for prediction of agricultural products prices

abstract autoregressive integrated moving average (arima) has been one of the widely used linear models in time series forecasting during the past three decades. recent studies revealed the superiority of artificial neural network (ann) over traditional linear models in forecasting. but neither arima nor anns can be adequate in modeling and forecasting time series since the first model cannot deal with nonlinear relationships and the latter one is not able to handle both linear and nonlinear patterns simultaneously. hence by combining arima with ann and designing the hybrid model, data relationship can be modeled more accurately. in this research, a hybrid of arima and ann models is designed and its prediction performance is compared with those of competing models. forecasting performance is examined using common criteria such as mse, rmse and mad. also the significance of any difference between these measures is tested through application of granger and newbold statistic. forecasting results for world wheat price data indicates that combined model significantly improves accuracy achieved by separate models.