Basic Informations

C.V

I am Asmaa Ahmed Abu Hamed, working as a Teaching Assistant at the Faculty of Politics and Economics at Beni-Suef University. I have a strong interest in statistics and scientific research, and I am always eager to enhance my research and academic skills. I have obtained a master's degree in statistics and aspire to continue scientific research and contribute to the development of this field.

Master Title

some proposed estimators for gamma regression model

Master Abstract

Regression models are vital statistical tools for examining relationships between a dependent variable and multiple independent variables. While linear regression is standard for continuous, normally distributed dependent variables, gamma regression is better suited for positive dependent variables following a gamma distribution. However, both models face challenges from multicollinearity, where high correlations among independent variables reduce the accuracy of parameter estimates and increase variance. This study introduces a new class of two-step shrinkage estimator, denoted as ß ^_GTSS, to address multicollinearity within the gamma regression model. The proposed estimator incorporates shrinkage techniques to enhance estimation accuracy, improve parameter precision, and effectively mitigate the adverse effects of multicollinearity. Its performance was evaluated through a Monte Carlo simulation study under varying conditions, including levels of correlation between independent variables and sample sizes. Comparisons were made against traditional estimation methods, such as maximum likelihood (ML) estimator, ridge estimator, Liu estimator, and the two-parameter estimator, using the mean squared error (MSE) criterion. The results demonstrated the superiority of the new estimator, particularly under severe multicollinearity, as it achieved the lowest MSE. Additionally, a real-world dataset analysis validated the practical effectiveness of the proposed estimator, showing substantial reductions in MSE. These findings highlight the estimator's potential to offer improved precision in parameter estimation, addressing a critical need for practitioners in this field. Keywords: Biased estimators; gamma regression model; Liu estimator; maximum likelihood estimator; mean square error; Monte Carlo simulation study; multicollinearity; new class of two-step shrinkage estimator; Ridge regression; two – parameter estimator; variance inflation factor.

PHD Title

000

PHD Abstract

000