PENGELOMPOKKAN DATA PENYAKIT DENGAN ALGORITMA K-MEDOIDS CLUSTERING PADA PUSKESMAS PLUS PERBAUNGAN

Abstract

Puskesmas Plus Perbaungan is a primary health facility that faces an increasing number of patients, especially those with hypertension. Manual data management makes it difficult to identify disease patterns for proper decision making. This study applies the K-Medoids
clustering algorithm to group hypertensive patients based on systolic and diastolic blood pressure, age, and activity. As a result, patients were divided into four clusters:
prehypertension (18-34 years old, alcohol), stage 1 hypertension (over 53 years old, smoking), stage 2 hypertension (40-52 years old, smoking), and normal blood pressure (35-
39 years old, alcohol). It is recommended that Puskesmas Plus Perbaungan increase
education about the risk of hypertension, conduct regular evaluations of clustering results, and consider other clustering methods to improve accuracy.