Comparative Evaluation of Machine Learning Models for PolyCystic Ovary Syndrome Prediction Using Phenotypic and Lifestyle Features

Poly Cystic Ovarian Syndrome is a prevalent endocrine disease among women of childbearing age. Its symptoms differs, and the diseases often coexist with other medical conditions, making diagnosis difficult. This study assesses eight machine learning models for PCOS prediction using a few lifestyle and phenotypic characteristics. To train and test eight machine learning models, standardized data is utilized. Performance of each model is compared using different Evaluation metrics. According to the evaluation, the best accuracy (0.83) and balanced performance are provided by the Support Vector Machine, k-Nearest Neighbor, and AdaBoost models. With self-as their appropriateness for early PCOS screening. This strategy might encourage early intervention and lessen long-term health issues.