PCOS data

This dataset integrates clinical and behavioral data from two sources: (1) a publicly available PCOS dataset from Kaggle and (2) a real-time survey conducted among 1,038 women aged 10\u201345 years from diverse backgrounds. The combined dataset contains 267 anonymized entries and includes 22 clinical and lifestyle features such as menstrual cycle regularity, acne, hair growth patterns, physical activity, and dietary habits. These variables were used to train and evaluate a deep learning model for the early detection of Polycystic Ovary Syndrome (PCOS), which achieved a test accuracy of 95%. The real-world survey data helped enhance the model\u2019s generalizability, while the Kaggle dataset provided a benchmark for initial training. All real-world data were collected with informed consent, and personally identifiable information was excluded. This dataset supports researchers working on AI-driven women\u2019s health diagnostics, particularly menstrual health and endocrine disorders.