مقارنة بعض طرائق تقدير معلمات الانموذج المختلط من الرتبة الاولى باستخدام المحاكاة == Comparison Some Parameters Estimation Methods For Mixed Model of Low Order Using Simulation
Author name:
لمياء محمد علي البدراني
Supervisor name:
لميعة باقر جواد الجواد
General topic:
Administration and Economics
Specific topic:
Statistics
Degree:
Doctorate
University:
University of Baghdad - Faculty Of Administration And Economics - Department Of Statistics
Language:
Arabic
University location:
Baghdad
First pages:
07T3980 - p.pdf
Abstract:
يهدف البحث الى دراسة طرائق تقدير معلمات الانموذج المختلط (الانحدار الذاتي - الوسط المتحرك) (Stationary Autoregressive - Moving Average Model) المستقر من الرتبة الاولى ARMA(1,1) مع دراسة الحالات الخاصة له، وهي : ARMA(1,0), ARMA(0,1) دراسة نظرية وتجريبية با | This thesis aims to studying the parameters estimation methods of the stationary mixed model (autoregressive - moving average) of low order ARMA(1,1) with the special cases of it which are : ARMA(1,0) and ARMA(0,1) in regard to time domain analysis in univariate time series.Using Exact maximum likelihood estimation methods (EML) and the approximating methods : backforecasting (BF) and Conditional least square (CLS) beside Moment method (M.M) with suggested conditional method (SC) and alternative method for moment method of ARMA(0,1) model. Driven some estimators for some special model.A comparison is done among the different methods by using criteria : mean square error (MSE) , mean absolute percentage error (MAPE) and average absolute error (AAE) and using several simulation experiments ,and iterating each experiments(1000) times, The results it found that the (EML) is better and more efficient than others. From the more notice conclusions that the relation between the sign and moment estimator for ARMA(1,1) model that is : when the sign is positive means the root gives invertable model and when the sign is negative means the root gives invertable model. The thesis consisted of four chapters, The first chapter contained introduction and literature review. The second chapter discussed the deferent estimation methods, The third chapter contained the experimental part using simulation for different sample sizes of series. It had been arrived at some conclusions and recommendations were consisting the fourth chapter
