عنوان مقاله [English]
نویسندگان [English]چکیده [English]
Often, the nature of many real life processes, especially in management and business fields are nonlinear. Forecasting the behavior of these processes requires accurate and effective forecasting tools. Shortages of such processes are removable by artificial neural network as an important modeling tool in business forecasting problems. In a comparing analyze, this paper shows the excellent performance of neural network in forecasting nonlinear processes rather than other forecasting models. For this, production, import and import value (dollar) data, related to wood industry of Iran, from 1961 to 2007 are studied. First, applying this data to neural network model and nonlinear models obtained from MATLAB software, the Iran wood industry was forecasted and then based on MAPE1, yielded outcomes from both models compared. Study findings show that in all cases neural network has more successful performance than models from MATLAB.