The Master’s student, Ikram Hakeem Ubais, successfully defended her thesis titled ‘Design and Implementation of Minkowski Features Selection and Aggregation for Machine Learning Algorithms’ at the College of Computer Science and Information Technology, University of Al-Qadisiyah. The research was conducted under the supervision of Assistant Professor Dr. Dhiah Al-Shammary.

The study aims to develop an aggregation algorithm capable of efficiently collecting dataset records into two groups while maximizing feature coherence. Integrating particle swarm optimization techniques enhances cluster accuracy by selecting highly ideal features. Intelligent feature selection represents an advanced stage in machine learning and innovative computer applications, reducing the number of required features for accurate classification. The study specifically focused on analyzing the best factors for feature selection in brain electroencephalogram data from the University of Bonn. Additionally, a comparison is made between the proposed system, other optimization techniques, and feature selection methods.

This study proposed two feature selection models. The first model introduces a novel approach for feature selection in machine learning using Minkowski similarity. Given the increasing complexity of high-dimensional medical datasets, effective feature selection methods are required for early disease detection and public health protection. The second model suggests an optimal aggregation approach using a particle swarm optimization tool based on Minkowski similarity. Experimental results demonstrated that the proposed model outperforms most optimization techniques in terms of accuracy metrics for the analysis and classification of brain electroencephalogram signals. Notably, the proposed model achieves an accuracy of up to 100%. These results serve as a significant incentive for experts in medical specialties, as they can benefit from a tool simplifying the identification and diagnosis of brain disorders.

The thesis was accepted with an excellent .