HARLPSO: Particle Swarm Optimization Based on Hungry Areas Random Learning for Medical Data Diagnosis

Medical problems are complex. In medical diagnosis, the most important thing is the accuracy of disease judgment. Due to the noise and redundant features in medical data, it is necessary to remove redundant features to improve the performance of the model. The particle swarm optimization algorithm (PSO) has a simple structure and can achieve an appropriate balance between exploration and exploitation during the search process. In the middle and later stages of the iteration of the PSO, the diversity of the population is lost. Therefore, this study proposes the Hunger Area Random Learning (HARL) strategy to open up new search spaces near the population. Experimental results show that the HARLPSO algorithm proposed in this paper outperforms existing feature selection algorithms on most medical datasets. This achievement provides a novel and efficient feature selection strategy and opens up new avenues for the analysis and diagnosis of biomedical data.