Discussion of the Diploma Research/ master’s thesis in the College of Computer Science and Mathematics -Department of (   Department of Statistics and Informatics   ) entitled:

supervised by  Dr.  Alaa Abdel Sattar Daoud

The thesis addressed       . The thesis addressed the use of wavelet analysis to improve penal regression methods and variable selection, focusing on the problem of outliers and errors with heavy-tailed distributions

The study addressed     . It touched on a set of theoretical and applied topics (linear regression analysis, outliers and influential observations, heavy-tailed distributions, selection of variables, penal regression methods, wavelet transforms, wavelet downscaling, the proposed methodology, integrating the wavelet technique with penal regression methods, and a simulation study. The thesis used Monte Carlo simulation to compare the performance of different methods under multiple conditions, such as differences in sample size, error distribution, contamination ratio, and presence of outliers

The study aims to  . The thesis aimed to employ wavelet transformation technology to improve the performance of fractional regression methods and variable selection, by smoothing the data and reducing the impact of noise, outliers, and errors with heavy-tailed distributions, and then integrating it with LASSO and SCAD methods and their robust counterparts (LAD-LASSO and LAD-SCAD), with the aim of improving the accuracy of estimating model parameters, increasing the efficiency of selecting important variables, increasing prediction accuracy, and evaluating the performance of these methods using simulation data and real data of thalassemia patients

The discussion committee consists of

Dr.   (Chairman)   Bashar Abdel Aziz Majeed

Dr.   (Member)    Osama Bashir Shukr

Dr .  (Member)    Mahmoud Mohamed Taher

Dr.  (Member and Supervisor)  Alaa Abdel Sattar Daoud