Swarm Optimization Feature Selection Algorithm for Real-Valued Information System under Mutual Information

This paper focuses on the measurement advantage of mutual information on feature correlation, combined with the global search ability of particle swarm optimization algorithm, and carries out the research on particle swarm optimization feature selection algorithm for real-valued information system under mutual information. First, mutual information and particle swarm optimization algorithm are introduced, and the evaluation model of real-valued information system based on mutual information is constructed. Then, through the fusion of mutual information gain calculation and particle swarm optimization iterative strategy, a particle swarm optimization feature selection algorithm for real-valued information system is proposed, which can efficiently identify redundant features and retain features with high contribution to classification or decision-making. Experimental results show that the proposed algorithm has excellent performance in feature selection.