Empowering Early Detection: A Web-Based Machine Learning Approach for PCOS Prediction
Pages:22-28
Poonam Musmade, Sachin Rajas, S. M. Khairnar, S.V. Rupanar
Polycystic Ovary Syndrome (PCOS) is a prevalent hormonal disorder impacting many women globally, leading to a range of complications from menstrual irregularities to infertility. The rise in PCOS incidence highlights the critical need for early detection and effective management strategies. This study explores a web-based machine learning approach to predict PCOS using a dataset of 541 patient records. Various machine learning models, including Logistic Regression (LR), Decision Tree (DT), Ada Boost (AB), Random Forest (RF), and Support Vector Machine (SVM), are employed to uncover patterns and predict PCOS. Feature selection is performed using the Mutual Information model, resulting in the highest accuracy of 94% achieved by both AB and RF models. The integration of machine learning techniques into a user-friendly web interface aims to enhance early PCOS detection.