Optimizing cancer classification using Krill Herd optimization with advanced feature selection techniques

Krill Herd Optimization (KHO) is a bio-inspired metaheuristic algorithm widely used in optimization problems. However, like many metaheuristic algorithms, KHO faces challenges related to convergence speed and exploration–exploitation balance. In this study, we explore the application and optimization of KHO for cancer classification by integrating Principal Component Analysis (PCA), Recursive Feature Elimination (RFE), and t -test for feature selection. The PCA technique is employed to reduce the dimensionality of the search space by extracting the most significant features, thus aiding KHO in navigating the solution landscape more efficiently. Subsequently, RFE is applied to further refine the feature set, ensuring that only the most relevant features are retained, thereby improving the convergence speed and solution quality. These techniques enhance the algorithm’s ability to identify the most relevant features, improving classification accuracy and computational efficiency. Our approach is evaluated on multiple cancer datasets, demonstrating improved performance over conventional methods. Specifically, our approach achieves an impressive accuracy score of 97% for PCA and 96% for RFE, showcasing its robustness and effectiveness in tackling high-dimensional optimization tasks. The results show that combining KHO with advanced feature selection techniques yields high accuracy in cancer classification, making it a promising approach for medical data analysis.