Application of Orangutan Optimization Algorithm for Feature Selection Problems

Feature selection plays a pivotal role in machine learning and data analysis by identifying the most relevant features to enhance model performance and reduce computational costs.This paper investigates the application of the Orangutan Optimization Algorithm (OOA), a bio-inspired metaheuristic, for solving feature selection problems.The performance of OOA is evaluated on three distinct feature selection tasks, demonstrating its effectiveness in identifying optimal feature subsets.Additionally, OOA is compared against eight state-of-the-art metaheuristic algorithms, including Genetic Algorithm (GA), Particle Swarm Optimization (PSO), Ant Colony Optimization (ACO), Grey Wolf Optimizer (GWO), and others.The results reveal that OOA consistently outperforms these algorithms in terms of classification accuracy, feature reduction rate, and convergence speed.The unique mechanisms of OOA, such as its memory-based decision-making and balanced explorationexploitation strategies, are highlighted as key contributors to its superior performance.This study establishes OOA as a robust and efficient tool for feature selection, with potential applications across various domains requiring dimensionality reduction and data preprocessing.