The College of Administration and Economics at the University of Baghdad discussed a master’s thesis in field of Statistics by the student (Narjes Bassem Khalaf ) and tagged with (Estimating logistical regression using nonparametric methods with application ) , Under supervision of ( Prof. Dr. leqaa Ali Mohammed)
The regression model is one of the most important statistical methods in analyzing data and describing the relationship between variables, so if we have a data set in which the response variable (Y) is measured in more than one way, it is quantitative or binary (0, 1) (healing, disease) (death, life), then there the logistic model is the optimal and most appropriate model, but this model varies according to the nature of the illustrative variables, if they are of a certain distribution and have a linear relationship with the dependent variable, the parametric logistic regression model was more appropriate. It is often the lack of data to these strict conditions, which drives them to search for another model represented by the nonparametric logistic regression model through the use of non-parametric estimation method represented by the methods of (kernel), slice (spline), Local Scoring Algorithm (LSA), Local Likelihood Logit Algorithm (LL) and mixed method (LBS), through the experimental side we concluded that the best estimation methods are the (LBS) method based on the comparison criterion Mean squares of error MSE, through the applied study that was carried out By studying lymphocytic leukemia, we found that when employing the best.
