ABSTRACT
Introduction: Metabolic syndrome is a group of risk factors for developing cardiovascular diseases and diabetes in an individual. The presence of various signs and symptoms makes the diagnosis of this disease difficult. Data mining can provide clinical data analysis of patients for medical decision-makings. The purpose of this study was to provide a model for increasing the predictive accuracy of metabolic syndrome
Method: In this applied-descriptive study, the medical records of 1499 patients with metabolic syndrome with 15 characteristics were investigated. Patients' information is collected from the standard database of Yazd Shohada-ye kargar Hospital. Each patient was followed for at least one year. In this paper, GBC algorithm was used to optimize the results of KNN data mining algorithm to predict and diagnose metabolic syndrome, and a new model was presented
Results: Based on the objective function to predict the increase of blood lipids in the proposed method, gray wolf algorithms, particle swarm and genetics were used to improve the performance of the KNN algorithm. The analyses show that the proposed model with the precision accuracy of 0.921 has a greater accuracy compared tofuzzy methods, backup vector machine, tree decomposition and neural network
Conclusion: Search in medical databases for the purpose of obtaining knowledge and information to predict, diagnose, and decision making are some applications of data mining in medicine. Hereditary algorithms can be used to optimize data mining techniques. The prediction and proper diagnosis of metabolic syndrome by using artificial intelligence and machine learning increases the chance of successful treatment