A high performance implementation of HDBSCAN clustering.
This project presents a comparative machine learning framework for heart attack risk prediction using Python. It includes data preprocessing, feature engineering, SMOTE for class balancing, cross validation, and performance evaluation of multiple ML models.