Polycystic Ovary Syndrome Prediction ML

Polycystic Ovary Syndrome (PCOS) is a significant health risk for women during their reproductive years.The disorder is characterized mainly by higher levels of male hormones and androgens, which cause the formation of fluid-filled follicles in the ovaries, preventing regular egg release.PCOS can result in difficulties such as miscarriage, infertility, and pregnancy troubles.According to the most recent information, about 31.3% of women in Asia have PCOS, and sadly, 69% to 70% of these instances go misdiagnosed.Recognizing the critical need for research to enable early detection and intervention to prevent serious PCOS effects, our primary research objective is to develop a predictive model for PCOS implementing advanced machine learning techniques.To establish our predictive models, we use a collection of clinical and physical information from women.We provide a novel feature selection method based on an optimized chi-squared (CS-PCOS) mechanism to improve accuracy and efficacy.The Gaussian Naive Bayes (GNB) method emerges as the top-performing model, overcoming other machine learning models and state-of-the-art studies, due to the novel CS-PCOS feature selection technique.GNB provides exceptional results with 100% accuracy, precision, recall, and F1scores while requiring only 0.002 seconds of calculating time.Our results highlight the importance of various dataset features such as , waist-hip ratio , prolactin (PRL), systolic and diastolic blood pressure, thyroidstimulating hormone (TSH), relative risk of breaths (RR-breaths), and pregnancy in PCOS prediction.Using the GNB algorithm and these critical criteria, our work intends to aid the medical community in early PCOS detection, thereby reducing miscarriage occurrences and enabling earlier management for women suffering from this disorder.