Identifikasi Pola Risiko Pasien Polycystic Ovary Syndrome (PCOS) Menggunakan Algoritma K-Means

Abstract

Polycystic Ovary Syndrome (PCOS) is a complex endocrine disorder commonly affecting women of reproductive age and associated with reproductive and metabolic conditions. Therefore, a systematic data analysis approach is needed to identify PCOS risk patterns. This study aims to cluster PCOS patient data using the K-Means algorithm and analyze the resulting risk patterns. The data were obtained from a public Kaggle dataset consisting of 2,539 patient records, including demographic, hormonal, metabolic, reproductive, and lifestyle variables. The research methodology follows the Knowledge Discovery in Databases (KDD) process, which comprises data selection, preprocessing, transformation, data mining, evaluation, and interpretation. The K-Means algorithm was applied to group the data into three categories: low, moderate, and high risk. The results indicate that the clustering process produces groups with distinct characteristics based on data similarities. The identified patterns include characteristics such as Body Mass Index (BMI), menstrual cycle irregularities, and hormonal parameters. These findings provide an overview of PCOS patient characteristic patterns that may serve as supporting information for healthcare professionals in understanding patient data and improving public awareness of PCOS prevention and management.