عنوان مقاله [English]
نویسندگان [English]چکیده [English]
This paper presents a new model for forecasting crude oil prices. The model is a combination of wavelet transformation with the ARMAX, Harmonic regression Holt-Winters and models. The study applies this model to forecasting the time series data of crude oil. The time series data of oil prices are decomposed by applying wavelet transformation to the three series; trend series, seasonality series and high frequency (fluctuations) series. The study then proceeds to apply the related models to forecast each series and finally to achieve the final forecasting, combines the forecasted time series with each other. By comparing the resulted forecasts from the proposed model with ARMA model, it is indicated that the using model in this thesis has better performance and accuracy.