الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 137 |  |  |
| | | from | 137 | | |
Abstract
The study and analysis of time series are one of the essential pillars in mathematical statistics, including studies of predicting the behavior of the time series in the future, which has many different applications in other sciences such as physics, medicine, astronomy, and economics. This thesis aims to create forecasting systems that give superior performance results by overcoming obstacles in univariate time series by introducing different strategies. The research consists of five chapters as follows:
First Chapter: Includes the problem statement the research Motivation, the research Objectives and Contributions and the description of the structure of this dissertation
Second Chapter: Includes the literature review, assessment of the literature in the field of trade rate prediction known and monetary prediction mainly using EMD, the overview of foreign exchange rate prediction and the overview of the hybrid models that incorporate the qualities of a couple of traditional models to get a higher gauging exactness.
Third Chapter: In this chapter, we proposed two models that integrate SEMD and ANN and EMD and ANN for improving the weakness of ANN. SEMD and EMD are an adaptive technique that shifts the non-stationary and non-linear time series data till it becomes stationary. In first stage, the data is decomposed into a smaller set of Intrinsic Mode Functions (IMFs) and residuals using EMD and SEMD. In the next stage, IMFs and residue are taken as the inputs for the ANN model. The methodology was compared with EMD-ANN and SEMD-ANN models. The results suggest that the SEMD-ANN is better than EMD-ANN.
Fourth Chapter: In this chapter, we proposed and illustrated statistical empirical mode decomposition based on neural network learning paradigm (SEMD-ANN) for forecasting Egypt stock market. By the criteria of some statistic loss functions, SEMD-ANN outperforms Holt-winters family model, EMD based on neural network (EMD-ANN) and SEMD and neural network (SEMD-ANN) in improving forecast accuracy.
Fifth Chapter: In this chapter, we conclude our research works that are done in this dissertation and describe our future work for improving the accuracy of the proposed forecasting time series techniques.