Development and Analysis of Machine Learning Models for Polycystic Ovary Syndrome Detection

Polycystic Ovary Syndrome (PCOS) is a common hormonal disorder that affects individuals with ovaries, particularly those of reproductive age. Applying machine learning (ML) to detect PCOS can offer several advantages, including early diagnosis, personalized treatment, and improved patient outcomes. This paper introduces a comprehensive approach to predict PCOS using ML techniques. The proposed model incorporates Synthetic Minority Over-sampling Technique (SMOTE) to address the class imbalance, Mutual Information (MI) based feature selection for interpretability, and an automatic threshold for optimized decision-making. The research evaluates ten ML models, including Linear Regression, Random Forest, Bayesian Ridge, Support Vector Machine, K-Neighbors Classifier, SGD Classifier, MLP Classifier, Logistic Regression, Gaussian Naive Bayes, and Gradient Boosting Classifier. The Random Forest classifier emerges as the top-performing model, achieving an accuracy of 92%. This highlights the success of the synergistic combination of advanced pre-processing techniques and robust classification algorithms. The results demonstrate the efficacy of the proposed methodology in accurate PCOS prediction, emphasizing the interplay between pre-processing strategies and powerful classification algorithms.