Application of XGBoost Model in the Field of Diabetes Prediction

Abstract

Diabetes is a metabolic disorder that threatens people's health, and standardized screening is an important way to diagnose and treat it early. It is low cost and high efficiency to screen through data, therefore, to predict diabetes early has become crucial. Diabetics were taken as the research subject in this paper, and XGBoost algorithm was used to process the patient's data from physical examination, so a model for predicting diabetes was established to predict the blood glucose level of patients and to explore the application of XGBoost model in the field of diabetes prediction. The experimental results have been shown that the mean square error of the sample using this model has been just 0.0598, and it have been verified that the prediction error of the model is small and the accuracy is high, which will soon provide a good means for the pre-screening and clinical prediction of diabetes.