A Comprehensive Survey on Dimension Reduction for Medical Big Data Using Optimization Algorithms

Abstract

This paper surveys dimension reduction techniques in medical big data using optimization algorithms to address challenges like computational inefficiency, overfitting, and inter-pretability in high-dimensional datasets. As medical data from sources like electronic health records, genomics, and imaging grow, efficient processing is essential for personalized healthcare. The paper explores feature extraction (PCA, LDA) and feature selection methods, emphasizing metaheuristic algorithms like Genetic Algorithms (GA), Particle Swarm Optimization (PSO), and Ant Colony Optimization (ACO). These algorithms enhance machine learning model accuracy by selecting relevant features, reducing computational costs, and handling nonlinear relationships in medical data. Applications in diagnosis, treatment prediction, and disease classification are discussed. Future research aims to integrate various optimization strategies and deep learning for more effective dimensionality reduction in healthcare.