Academic/Professional Position (Other)
Assistant Professor
Presentation Type
Oral Presentation
Discipline Track
Biomedical ENGR/Technology/Computation
Abstract Type
Research/Clinical
Abstract
Background: The system's performance may be impacted by the high-dimensional feature dataset, attributed to redundant, non-informative, or irrelevant features, commonly referred to as noise. To mitigate inefficiency and suboptimal performance, our goal is to identify the optimal and minimal set of features capable of representing the entire dataset. Consequently, the Feature Selector (Fs) serves as an operator, transforming an m-dimensional feature set into an n-dimensional feature set. This process aims to generate a filtered dataset with reduced dimensions, enhancing the algorithm's efficiency.
Methods: This paper introduces an innovative feature selection approach utilizing a genetic algorithm with an ensemble crossover operation to generate the most robust offsprings, i.e., feature set solutions. The selection process is driven by computed minimum objective function values (OFVs). Experimental evaluations were conducted across diverse datasets sourced from various repositories, all processed through a standardized classifier. A comparative analysis was performed, contrasting the proposed feature selection system with various traditional counterparts.
Results: Our innovative approach yielded superior results compared to conventional feature selection techniques in terms of accuracy and the reduction in the number of features. This holds significant promise as a valuable tool for the diagnosis of lung cancer and pancreatic cancer.
Conclusion: In addition to surpassing traditional feature selection techniques in accuracy and feature reduction, our novel approach holds great potential across diverse fields, with a particularly promising impact in the medical domain for early diagnostics. This advancement could contribute significantly to the timely and effective identification of medical conditions, enhancing the overall capabilities of diagnostic processes.
Recommended Citation
Khan, Shabia Shabir; Kawoosa, Majid Shafi; Bannerjee, Bonny; Chauhan, Subhash C.; and Khan, Sheema, "Revolutionizing Feature Selection: A Breakthrough Approach for Enhanced Accuracy and Reduced Dimensions, with Implications for Early Medical Diagnostics" (2024). Research Symposium. 24.
https://scholarworks.utrgv.edu/somrs/2024/talks/24
Included in
Computational Engineering Commons, Endocrine System Diseases Commons, Neoplasms Commons, Other Computer Engineering Commons, Respiratory Tract Diseases Commons
Revolutionizing Feature Selection: A Breakthrough Approach for Enhanced Accuracy and Reduced Dimensions, with Implications for Early Medical Diagnostics
Background: The system's performance may be impacted by the high-dimensional feature dataset, attributed to redundant, non-informative, or irrelevant features, commonly referred to as noise. To mitigate inefficiency and suboptimal performance, our goal is to identify the optimal and minimal set of features capable of representing the entire dataset. Consequently, the Feature Selector (Fs) serves as an operator, transforming an m-dimensional feature set into an n-dimensional feature set. This process aims to generate a filtered dataset with reduced dimensions, enhancing the algorithm's efficiency.
Methods: This paper introduces an innovative feature selection approach utilizing a genetic algorithm with an ensemble crossover operation to generate the most robust offsprings, i.e., feature set solutions. The selection process is driven by computed minimum objective function values (OFVs). Experimental evaluations were conducted across diverse datasets sourced from various repositories, all processed through a standardized classifier. A comparative analysis was performed, contrasting the proposed feature selection system with various traditional counterparts.
Results: Our innovative approach yielded superior results compared to conventional feature selection techniques in terms of accuracy and the reduction in the number of features. This holds significant promise as a valuable tool for the diagnosis of lung cancer and pancreatic cancer.
Conclusion: In addition to surpassing traditional feature selection techniques in accuracy and feature reduction, our novel approach holds great potential across diverse fields, with a particularly promising impact in the medical domain for early diagnostics. This advancement could contribute significantly to the timely and effective identification of medical conditions, enhancing the overall capabilities of diagnostic processes.