The use of ARIMA, ANN and SVR models in time series hybridization with practical application | ||
| International Journal of Nonlinear Analysis and Applications | ||
| مقاله 8، دوره 14، شماره 3، خرداد 2023، صفحه 87-102 اصل مقاله (595.39 K) | ||
| نوع مقاله: Research Paper | ||
| شناسه دیجیتال (DOI): 10.22075/ijnaa.2022.7168 | ||
| نویسندگان | ||
| Noor Saleem Mohammed Ali؛ Firas Ahmed Mohammed* | ||
| College of Administration and Economics, Department of Statistics, University of Baghdad, Iraq | ||
| چکیده | ||
| Forecasting is one of the important topics in the analysis of time series, as the importance of forecasting in the economic field has emerged in order to achieve economic growth. Therefore, accurate forecasting of time series is one of the most important challenges that we seek to make the best decision, the aim of the research is to suggest employing hybrid models to predict daily crude oil prices. The hybrid model consists of integrating the linear component, which represents Box Jenkins models, and the non-linear component, which represents one of the methods of artificial intelligence, which is the artificial neural network (ANN), support vector regression (SVR) algorithm and it was shown that the proposed hybrid models in the prediction process when conducting simulations for the time series and for different sample sizes and when applying them on the daily crude oil price data, it was more efficient than the single models, as the comparison between the single models and the proposed hybrid models was done by means of the comparison scale, the mean square error (MSE), the results showed that the proposed hybrid models gave more accurate and efficient results, in addition to its ability to predict crude oil prices well. | ||
| کلیدواژهها | ||
| ARIMA؛ Artificial Neural Network؛ Time Series Forecasting؛ Support Vector Regression؛ Hybrid Model | ||
| مراجع | ||
|
| ||
|
آمار تعداد مشاهده مقاله: 17,499 تعداد دریافت فایل اصل مقاله: 10,936 |
||
| تعداد نشریات | 22 |
| تعداد شمارهها | 722 |
| تعداد مقالات | 10,383 |
| تعداد مشاهده مقاله | 72,839,621 |
| تعداد دریافت فایل اصل مقاله | 64,536,194 |