master’s thesis by ” Noor abd elmunm” Department of  Statistics and Informatics

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

supervised by  Dr. Omar Qusay jasim

The thesis addressed the problem of variable selection in high-dimensional data, specifically gene expression data associated with cancer classification.

 

The study focused on the application of the Performance-Weighted Resampling-based Combinatorial Variable Selection (PWRCVS) method, integrating selection stability, predictive performance, and variable importance while accounting for inter-variable correlations.

The thesis aimed to develop and evaluate the PWRCVS method for selecting a small, stable subset of variables—while maintaining strong classification performance—and to compare it with traditional selection methods using real-world high-dimensional data.

This thesis achieves the following Sustainable Development Goals:

1: Good Health and Well-being, through the analysis of cancer-related gene expression data.

2: Industry, Innovation, and Infrastructure, through the development of an innovative statistical method for variable selection.

3: Quality Education, through supporting scientific research and advancing knowledge in high-dimensional data analysis.

The discussion committee consists of

Dr. Bashar Abdul-Aziz Majeed(Chairman)

Dr.  Osama Bashir Shukr (Member)

Dr . Ban Nawzat Ahmed(Member)

Dr. Omar Qusay jasim  (Member and Supervisor)