Machine Learning-Driven PCOS Prediction for Early Detection and Tailored Interventions
B. Yamini, Venkata Ramana Kaneti, P Prema, C. Ambhika, M. Nalini, R. Sıva Subramanıan
Abstract
This PCOS is a hormonal disorder that leads to the overproduction of androgens, resulting in symptoms such as interrupted periods, ovarian follicles, excess body hair, weight gain, and infertility. Hormonal imbalances and abnormal male hormone production characterize it. The precise aetiology of Polycystic Ovary Syndrome is unknown, but insulin resistance and genetic factors may play a role. Due to its prevalence and potential long-term health effects, understanding and predicting PCOS is most important in healthcare. The study of PCOS in women aids in identifying the condition earlier and protecting women from life-threatening medical complications. This research uses ML algorithms to develop a novel predictive modelling strategy to identify individuals at risk of developing PCOS. In the field of reproductive health, ML has the potential to revolutionize healthcare by enhancing detection and prediction. Multiple ML models, such as LR, RF, SVM, NB, K NN, and XGBoost, were used to predict PCOS. The examination uses a PCOS dataset containing clinical, hormonal, and biological information from women with and without PCOS issues. The acquired experimental results are projected using various validity metrics, including precision, recall, accuracy and F1-Score. The outcome indicates that Machine Learning models have promising predictive ability, and the random forest model has a 90% accuracy rate, which is higher than any other model. PCOS research is essential for encouraging early diagnosis, effective treatment, and improved reproductive health outcomes for those affected. Using Machine Learning algorithms, our proposed method provides a promising approach to PCOS prediction, enabling physicians to rapidly identify at-risk patients and perform tailored therapies. By treating PCOS early, healthcare practitioners may help women with this complicated endocrine illness avoid problems and enhance their quality of life.