Leveraging Semi-Supervised Learning for Early Diagnosis of Polycystic Ovary Syndrome (PCOS)

Polycystic Ovary Syndrome (PCOS) is a prevalent hormonal disorder that can lead to serious health complications, including infertility, diabetes, and cardiovascular issues. Early detection is critical for effective management and treatment, but the challenge of limited labeled data persists. This study explores the use of Semi-Supervised Learning (SSL) techniques for PCOS detection, leveraging both labeled and unlabeled data to enhance the performance of machine learning models. Ten different machine learning classification algorithms, including Logistic Regression (LR), Decision Tree (DT), AdaBoost (AB), Random Forest (RF), and Support Vector Machine (SVM), were employed and compared in both SSL and Supervised Learning (SL) settings. The experimental results show that SSL models achieved accuracies ranging from 79.14% to 91.37%, which are comparable to, or even exceed, those of the SL models. Among the SSL models, AdaBoost and Random Forest achieved the highest accuracies of 91.37% and 89.21%, respectively. The SSL models also demonstrated competitive performance in terms of sensitivity, specificity, and AUC scores, proving their effectiveness in detecting PCOS with minimal labeled data. These findings highlight the potential of SSL as a promising alternative to traditional SL methods in PCOS diagnosis, offering a more efficient and reliable solution while reducing the reliance on large labeled datasets.