<![CDATA[This study aims to develop a web-based hypertension detection system using the K-Nearest Neighbors (KNN) algorithm and to analyze its accuracy in classifying hypertension status. The dataset was obtained from 447 patient medical records at RSKG Rasyida, consisting of eight variables: gender, age, systolic blood pressure, diastolic blood pressure, height, weight, body mass index (BMI), and hypertension status. The preprocessing stage involved three main steps—feature selection (age, systolic and diastolic blood pressure, BMI), data balancing using undersampling, and data normalization through the Min-Max method—resulting in 425 balanced data samples with five hypertension categories. The web application includes modules for login, dashboard, data input, detection results, and detection history, and has been evaluated using black box testing. The best KNN performance was achieved at k = 13 with 92.94% accuracy, 94% precision, 93% recall, and 93% F1-score. These results indicate that the proposed system can accurately classify hypertension and serve as an effective, data-driven screening tool for healthcare professionals.]]>
