Hybridization of Meta-heuristic Algorithms with K-Means for Clustering Analysis: Case of Medical Datasets

K-Means clustering is commonly used for data clustering, but it suffers from limitations such as being prone to local optima and slow convergence, particularly when handling large medical files.The literature recommends employing metaheuristic algorithms in clustering studies to address these issues.This study aims to accurately diagnose diseases in four medical datasets (Dermatology, Diabetes, Parkinson's, and Thyroid) and increase the rate of correct diagnosis of diseases.We utilized optimization algorithms to assign weights to input parameters determining diseases in these datasets, thereby improving clustering performance.Our proposed model incorporates the Crow Search Algorithm, Tree Seed Algorithm, and Harris Hawks Optimization algorithms in a hybrid structure with K-Means.We conducted statistical evaluations using performance metrics.The study demonstrated that the hybrid Harris Hawks Optimization algorithm achieved the highest accuracy rate (97.19%) among the tested algorithms on the Dermatology dataset.The hybrid Crow Search Algorithm obtained a 96.29% accuracy rate on the Thyroid dataset, while the hybrid Tree Seed Algorithm achieved a 95.32% accuracy rate on the Dermatology dataset.This study offers significant benefits, including reduced staff workload, lower test costs, improved accuracy rates, and faster test results for detecting various diseases in medical datasets.