Optimization Algorithms for Feature Selection in Healthcare

Abstract

Feature selection is a crucial step in healthcare data analytics, as high-dimensional medical datasets often contain redundant and irrelevant information that can degrade model performance. This paper presents a framework based on Optimization Algorithms for Feature Selection in Healthcare, aiming to enhance predictive accuracy while reducing computational complexity. The proposed approach utilizes metaheuristic optimization techniques such as Genetic Algorithms, Particle Swarm Optimization, and other nature-inspired methods to identify optimal subsets of features from clinical datasets. These algorithms are integrated with machine learning models to improve classification performance and minimize overfitting. The framework also incorporates hybrid optimization strategies that combine global search capabilities with local refinement techniques for improved convergence. Experimental results demonstrate that the proposed method significantly enhances model accuracy, reduces feature dimensionality, and improves computational efficiency. Additionally, integration with prior research in ensemble learning and healthcare analytics strengthens system robustness and generalization. The study highlights the importance of optimization-driven feature selection in developing efficient, scalable, and reliable healthcare predictive systems.