AI-Powered Heart Disease Prediction App
Technical project
Summary
• Developed and deployed a Random Forest Classifier model, achieving 88% accuracy and ROC-AUC of 0.91 for heart disease risk prediction. • Engineered data preprocessing pipelines (feature scaling, encoding, and hyperparameter tuning) using Scikit-learn and GridSearchCV, improving reliability by 12%. • Deployed the model via Flask web application with real-time user input and dynamic results rendering using Jinja2 templates. • Integrated CI/CD workflow with GitHub and Render for automated deployment and version control for public access.