Welcome to the Heart Failure Prediction System – a machine learning-powered web application that predicts the likelihood of heart disease in patients using medical parameters.
🔗 Live App: heart-disease-app.vercel.app
This project helps in predicting the risk of heart failure using patient data. It uses a trained machine learning model and provides an intuitive web interface for real-time predictions.
Key features:
- Built with Python, Streamlit, and deployed using Vercel
- Predicts heart disease based on various clinical features
- Logistic Regression used as the core predictive model
- Clean and responsive UI for easy interaction
-
📚 Dataset Used: Heart Disease UCI dataset
-
🧹 Preprocessing:
- Handled missing/null values
- Encoded categorical variables like chest pain type, slope, etc.
- Scaled features for better model performance
-
🤖 Model: Logistic Regression
- Accuracy: ~86%
- Evaluated using confusion matrix, ROC-AUC score, and classification metrics
- Real-time heart disease prediction
- User-friendly input form
- Backend logic in Python
- Deployed seamlessly on Vercel for production use
├── Heart_Failure_Prediction.ipynb # Jupyter Notebook with full model pipeline
├── requirements.txt # Python dependencies
├── README.md # Project documentation
- Python
- Pandas, NumPy, Scikit-learn, Matplotlib, Seaborn
- Streamlit (as the web interface framework)
- Vercel (for frontend hosting)
This application was deployed using Vercel. The frontend is powered by Streamlit, and the entire app was exported and configured for Vercel deployment using a vercel.json config and proper structure to serve the Streamlit interface as a static app via Vercel’s serverless platform.