Early Prediction and Diagnosis of Hypertension Using Machine Learning Techniques

High blood pressure (HBP) is a major health problem worldwide, causing heart and blood vessel disease and may not have any symptoms in the early stages. In this study, machine learning models are compared namely XGB Classifier, LGBM Classifier, and Random Forest Classifier to predict hypertension using a dataset from Kaggle with 80% for training and 20% for testing. The accuracy, precision, recall, F1-score, and computational efficiency (training and prediction times) were used to evaluate the performance. All three models were able to reach 100% accuracy on the test set, where precision, recall and F1-score were 1.00 for both classes of Hypertension and Normal. The training time for XGB Classifier, LGBM Classifier and Random Forest Classifier were 0.0258s, 0.0214s and 0.1795s respectively and the prediction time were 0.0025s, 0.0019s, and 0.0077s respectively. The results point out that the evaluation of the model should not only rely on the accuracy but also on the computational efficiency and data features of the model. Establishing a foundation for early-prediction clinical support systems is important, and external validation and dataset diversification are still crucial in preparation for clinical deployment.