This research paper focuses on the application of unsupervised learning techniques for clustering medical data in order to predict the likelihood of polycystic ovary syndrome (PCOS) in women. The study explores the utilization of unsupervised learning algorithms, namely K-means and hierarchical agglomerative clustering, to create clusters from a dataset comprising various medical specifications of women. By employing these algorithms, the research aims to classify the data points into distinct groups, thereby facilitating the identification of patterns and correlations among different variables. The clustering approach offers a valuable means to uncover hidden relationships and dependencies within the dataset, aiding in the development of predictive models for PCOS likelihood in women. Through an analysis of the clustered data, this research endeavors to contribute to the field of medical diagnostics by providing insights into the potential early detection and intervention of PCOS in women.