According to the importance of oil in economy of the world, different models have been developed for formulating the behavior of oil price. It is due to the fact that the models established based on the new techniques are more reliable and accurate on account of their ability in taking into account either linear or non-linear structures involved in the process of oil pricing. Auto-Regressive Integrated Moving Average (ARIMA) method is one of the most common time series models applied in forecasting over the resent decades. These researches indicate that two major limitations are found in past models: (1) ARIMA approach assume that there is a linear relationship between the future values of a time series with current and past values as well as a white noise; so that estimations obtained by ARIMA cannot be appropriate for modeling the nonlinear problems; and (2) a large number of historical data are required to satisfy the results. On the other hand, adaptive neuro-fuzzy inference system (ANFIS) is a powerful tool for modeling the non-linear structures. In this paper, ARIMA and ANFIS are used to overcome the limitations of conventional models, thus obtaining more accurate results. Empirical results of oil price forecasting indicate that the proposed model outperforms other methods and exhibits the accuracy of oil price forecasting is improved; so that, the proposed model can be applied as a proper option to forecast financial time series.
کلید واژگان :oil price
ارزش ریالی : 300000 ریال
با پرداخت الکترونیک