Researchers have developed a Random Forest machine learning model to predict hypertension in patients with type 2 diabetes that combines clinical data, lifestyle information and socioeconomic factors. The study included 900 patients divided into groups for training, testing and external validation of the model. The model achieved high accuracy with an AUC value of 0.89 in internal testing and 0.83 in external validation, which confirms its reliability. Key predictors of hypertension were alcohol consumption, triglyceride level, duration of diabetes, type of health insurance, fasting blood glucose, kidney function, and exercise frequency. The model was explained using SHAP analysis, which provides a clear understanding of how individual factors affect the risk of hypertension. The researchers concluded that this validated model can help doctors in early identification of patients at high risk of hypertension and in personalized prevention of the disease in clinical practice.