Artificial Intelligence-Based Proactive Diagnosis of PCOS and PCOD through Clinical and Hormonal Measures

Polycystic Ovary Syndrome and Polycystic Ovary Disorder are common female hormonal disorder which cause irregular period, infertility, obesity, and other metabolic problems. Hence, the management of these disorders and early prediction is of great importance. The current project implies the use of data- driven methodology to identify and forecast PCOS/PCOD using machine learning algorithms, including Logistic Regression and Random Forest. The data set includes the varied scope of clinical and lifestyle characteristics, such as Age, BMI, Pulse rate, Cycle length, FSH, LH, AMH, PRL, and TSH levels and ultrasound measurements. In this project, various preprocessing practices are performed, and the main attention is paid to data normalization, categorical attributes encoding, and missing values. In the given project, models are trained and evaluated in terms of such performance measures as accuracy, precision, recall, and F1- score. Random Forest was found to be better than another model as it produced the highest accuracy. Thus, one can possibly indicate that this model is the most dependable and could be used to deal with nonlinear interactions between features. In addition, the model does not only categorize whether a particular individual is a PCOS or PCOD patient but also identifies the most potent attributes, which result in the occurrence of the disorder. As such, it is an effective decision support system that can be used to assist clinicians in making appropriate decisions.