استخدام التقليص المويجي في كشف و معالجة القيم الشاذة في تحليل الانحدار الخطي و مقارنته مع بعض الطرائق الحصينة

Abstract EN

In this research the (Waveshrinkage) approach is used , to treat the Outliers Values during fitting The Simple Linear Regression Model by using some wavelets on the account of Discrete Wavelet Transformation (DWT) with one of the Threshold kinds To obtain a modified coefficients for (DWT) special dependent variable and depending on estimation of Linear Regression Model, and account some statistic measurements, mean squares error, Determination Coefficient and F-test (MSE,R2,F) and compared to Some Robust Method and it’s a way of Least Median Square Method (LMS) and Least Trimmed Squares Method (LTS) during two experiments ,one of them depends on data which present strength point of dependent variable and solid hardness's of independent variable while the second one presents the amount of sugar which is dependent variable and the proportion of vegetable fats, cholesterol level in the blood, the patient's age and weight of the patient as independent variables for (50) patients of diabetes in Layla Qasim Hospital in Erbil city by depending on statistics instrument of programmed (S-Plus), (SPSS), in addition to usage of (MATLAB) Language.