<![CDATA[Hypertension is one of the most prevalent non-communicable diseases and a major public health problem because it increases the risk of cardiovascular disease and other serious complications if not detected early. Therefore, a system capable of classifying blood pressure accurately and consistently may support preliminary assessment. This study aims to develop and evaluate a hypertension classification system based on the Fuzzy K-Nearest Neighbor (FKNN) method using systolic and diastolic blood pressure parameters. The system was developed using 814 training data and 202 testing data, classified into three categories: Hypertension, Prehypertension, and Normal. The classification process consists of data normalization, Euclidean distance calculation between testing and training data, determination of the K nearest neighbors, calculation of fuzzy membership degrees for each class, and assignment of the class with the highest membership value. The system was evaluated using K values ranging from 3 to 30 with a fuzziness parameter (m) of 2 to analyze the effect of the number of neighbors on classification performance and identify the optimal K value. System performance was assessed using accuracy, precision, recall, and F1-score metrics. The experimental results showed that the FKNN method achieved high classification performance, with the highest accuracy of 98.02% at K = 3, followed by 97.03% at K = 5 and 96.53% at both K = 9 and K = 15. The precision, recall, and F1-score values across the three classes also indicated good classification performance. These findings suggest that the FKNN method has potential for classifying blood pressure based on systolic and diastolic measurements, with K = 3 yielding the highest accuracy among the tested parameter settings. However, the evaluation was limited to the available dataset and did not include external validation, assessment by clinical professionals, or prospective evaluation within clinical workflows. Therefore, the developed system should be regarded as a preliminary classification prototype, and its clinical applicability and operational suitability as a decision-support tool require further investigation.]]>
Copyrights © 2026
