الفهرس 
PreliminariesOnly 14 pages are availabe for public view
 |  | from | 198 |  |  |
| | | from | 198 | | |
Abstract
The spectral analysis of time series with missing data is one of the most important problems faced by applied researchers whose data arise in the form of time series. When observing a time series at equal spaced intervals of time, it might be happen that the device being used to observe the series will miss an observation because of some random failure. We can see that in many branches of science, particularly in electrical engineering, physics, meteorology, marine science and economics. In this thesis we concerned with studying this case. The basic aims of the thesis can be summarized in, (1) Studying the spectral analysis of a strictly stationary r-vector valued discrete time series in the case where there are some randomly missing observations in disjoint and joint segments of observations. (2) Studying the statistical properties of each estimate (3) Studying some discrete time series models in the time domain by using the dynamical system approach. (4) Using numerical simulation to study theoretical results. (5) Studying some numerical applications of the theoretical results. (6) Studying the nonparametric spectral analysis of continuous time series with randomly missing observations. (7) Studying the statistical properties of each estimate. The thesis consists of three chapters. In Chapter 1, we introduce a survey about time series analysis methods and its developments. Basic definitions and theorems that used in the thesis are also introduced. The primary focus in Chapter 2 is on studying the spectral analysis of strictly stationary discrete time series in disjoint and joint segments of observations with missing values. We also study the dynamics of EXPAR models. Chapter 3 is devoted to study the nonparametric spectral analysis of a strictly stationary r-vector valued continuous time series with some randomly missing values.