<![CDATA[Polycystic Ovary Syndrome (PCOS) is a common endocrine disorder occurring in women of reproductive age, with a global prevalence ranging from 6% to 21%. Current management of PCOS remains limited to symptomatic treatment without addressing the root cause. This study aims to build an accurate predictive model for PCOS diagnosis in Indonesia by analyzing lifestyle and reproductive factors using machine learning algorithms, such as Logistic Regression and Random Forest.The research dataset consists of 541 patient records, which were divided into 80% for training and 20% for testing. The data was normalized using the Min-Max Scaler method, and class imbalance was handled using the SMOTE (Synthetic Minority Oversampling Technique) method. The models were validated using the K-Fold Cross-Validation method and evaluated based on accuracy, precision, recall, and F1-score.The results showed that Logistic Regression with SMOTE in predicting reproductive factors achieved the highest accuracy (82%), while Random Forest with SMOTE demonstrated more stable performance based on average accuracy, particularly for reproductive factors. ROC curve analysis also revealed that Logistic Regression with SMOTE in predicting reproductive factors achieved the highest AUC ($0.84$), making the Logistic Regression model superior in predicting the diagnosis compared to Random Forest. This study confirms that reproductive factors play a more dominant role in predicting PCOS compared to lifestyle factors. Utilizing machine learning algorithms can effectively predict PCOS to support management and prevention, as well as accelerate the early detection process of PCOS.]]>
