الفهرس 
مقدمة البحثيوجد فقط 14 صفحة متاحة للعرض العام
 |  | from | 143 |  |  |
| | | from | 143 | | |
المستخلص
Several time series models and Regression models are used in forecasting financial time series, but also several time series in economy haven’t constant mean and a lot of them reflect relative nonstationarity followed by high volatility which needs to extend time series analysis. Some important variables in macro economy show nonstationarity especially in finance; since means of the sample don’t show constant and there is obviously appearance of hetroskedasticity.
This research aims to identify the best forecasting model of stock prices to be more accurate regarding the problem of unstationarity of the time series mean and variance, by using combining some regression models as Support Vector Regression and Grey Model with time series GARCH Model by neural networks .
In this study comparison between GARCH model, Grey model, Support Vector Regression model, Support Vector Regression with Grey model, GARCH model with Grey model and combining Grey model and Support Vector Regression with GARCH model by neural networks has been done. This study reaches that using combining Grey model and Support Vector Regression model with GARCH model by neural networks is the best and the most accurate in forecasting stock prices, The applied study deals with a group of daily data of ORSCAM Telecome company. The research recommends extending the use of time series analysis as an effective tool in studying many financial variables and forecasting of it. Also the research emphasizes the importance of volatility in this kind of data not only as a variable has an important mean but also as a necessary explanatory variable in understanding the behavior of many variables in finance.