Revolutionizing Metabolic Syndrome Detection and Classification with State-of-the-Art Machine Learning Techniques

Metabolic syndrome (MS) is a medical disorder characterized by a number of metabolic abnormalities, such as hypertension, obesity, hyperglycemia, and hyperlipidemia. Improved cardiovascular disease-related health outcomes can be achieved with early MS risk prediction in middle-aged adults. To enable an early and accurate diagnosis, a multitude of machine learning algorithms have been developed to support the clinical diagnosis of metabolic syndrome. This work aimed to identify the optimal MS prediction model using the state-of-the-art machine learning methods. Using a dataset, we have implemented machine learning algorithms for the purpose of early diagnosis and prediction of MS. For MS prediction, feature selection strategies were combined with machine learning models such logistic regression, XGBoost, Adaboost, Random Forests, Decision Trees, Support Vector Machines, and Naïve Bayes. After analyzing the results of each classifier, it was revealed that the Adaboost, Catboost and Random Forest classifier had the greatest and most dependable accuracy of 83.5%, 82.3% and 83%.