Metabolic Syndrome Risk Forecasting on Elderly with ML Techniques

Abstract

Metabolic syndrome is a disorder that affects the overall function of the human body. It is manifested by elevated levels of cholesterol and triglycerides, a significant reduction in energy levels, weight gain with visceral fat deposition in the abdomen, and menstrual disorders while increasing the risk of cardiovascular disease, autoimmune diseases and diabetes. A public dataset is exploited to evaluate the metabolic syndrome (MetS) occurrence risk in the elderly using Machine Learning (ML) techniques concerning Accuracy, Recall and Area Under Curve (AUC). The stacking method achieved the best performance. Finally, our purpose is to identify subjects at risk and promote earlier intervention to avoid the future development of MetS.