تحديد افضل اسلوب تمهيدي حصين لتقدير انموذج انحدار لا معلمي مع تطبيق عملي == Determination of The Best Robust Smoothing Technique To Estimate A Nonparametric Regression Model With Practical Application
Author name:
غياث حميد مجيد
Supervisor name:
فارس طاھر حسن الكواز
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:
07T3929 - p.pdf
Abstract:
لقد شهدت طرائق تقدير نماذج الانحدار اللامعلمي في السنوات الماضية توسعا كبيرا، ويعزى ذلك التوسع الى توصل الباحثين الى صعوبة مواكبة الطرائق المعلمية المستعملة في تقدير نماذج الانحدار بسبب المرونة المطلوبة عند تحليل البيانات سواء كانت البيانات كمية ام نوعي | In the past many years, the methods of estimating nonparametric regression models have been witnessed a significant expansion, due to the researchers reach difficulties to cope with use of the parametric methods in estimating of regression methods because of the desired flexibility during data analysis whether being qualitative or quantitative. But this expansion in the field of estimating nonparametric regression models come up with the major development in the hardware and software of computers, which leads to development of nonparametric regression methods and one of the most important methods is smoothing methods.In this research, the study takes different types of smoothing methods which include Kernel Smoothing and Smoothing Splines and their significant role in estimation of nonparametric regression models to describe the relationship between the explanatory variables and response variable. The researcher employs the concept of Robustness in the smoothing methods that have been adopted in this research, by using three types of smoothing methods which are Local Polynomial Kernel (LPK), Penalized Spline (PS) and Spline Regression (SR) which ends up with robust smoothing methods which are, Robust LPK, Robust PS, and Robust SR.Many issues may face the researcher while smoothing nonparametric regression models, such as data includes outliers. This leads to use the robust methods when smoothing the nonparametric functions. To do this certain methods of robust nonparametric regression has been used instead of some suggested methods in case of outliers exists in data. Therefore, the aim of this dissertation is to estimate the nonparametric regression model using robust smoothing methods and making comparisons between different methods. The researcher took the most useful researches deals with this subject such as robust methods that use these methods to estimate nonparametric regression model, and apply them in two fields, experimental and practical, in experimental field we use the simulation technique to have constant data that simulate the real data that use it in practical field. In practical field real data has been taken from Iraqi Stock Market especially the data about Al - Khaliej Insurance Company, the response variable represents the Close Price for Stock and the elementary variable is the Trading Volume. Also, the researcher suggested some of the most important conclusion drawn from experimental and practical fields in addition to some recommendations that can be adopted in future studies
