IDENTIFIKASI FAKTOR UTAMA PENYEBAB SINDROM OVARIUN POLIKLISTIK (PCOS) MENGGUNAKAN ALGORITMA C4.5

Abstract

Polycystic ovary syndrome (PCOS) is a common hormonal disorder in women of reproductive age, characterized by ovarian dysfunction, high androgen levels and insulin resistance. According to WHO, 6-13% of women have PCOS, and up to 70% are undiagnosed. This study aims to predict the main factors causing PCOS by using the C4.5 algorithm based on clinical data attributes, such as BMI, Menstrual Irregularity, Testosterone Level(ng/dL), and Antral Follicle Count. It can be concluded that the Menstrual Irregularity attribute is the most dominant factor followed by BMI, Testosterone Level (ng/dL), and Antral Follicle Count. The developed model achieved 83% accuracy, 94% precision, and 78% recall, showing strong capability in identifying positive PCOS cases with minimal error rate. Comparison between the results of manual calculation using Excel and the automatic process through RapidMiner resulted in the same tree structure, thus confirming the credibility and consistency of the method used.