استعمال الطرائق الحصينة لتقدير معالم انموذج الاقتران لجداول التوافق بوجود خلايا شاذة مع التطبيق العملي

Abstract

Abstract In this research estimate the parameters of the model association in the case of the existence outliers cells are an important way to analyze the data to see any of the estimation methods classical and robustness is the best and most efficient when using M-Estimators using the functions (Huber, Tukey, Andrews) When using method (Kuhnt) for the detection of outliers in the Contingency Tables to the real data concerning the phenomenon of unemployment for a sample of (1834) unemployed, in Wasit City, and found three of outliers cells is equivalent to the proportion (10%) of the data, it turns out that robust M-Estimators, the function (Andrews) is the best among the functions used and this is proven by the standard differentiation (Mean Square Error MSE),