A machine learning application that predicts heart disease risk based on patient symptoms and diagnostics using the Cleveland Heart Disease Dataset from UCI.
This application uses two powerful machine learning algorithms:
- Random Forest: An ensemble learning method that operates by constructing multiple decision trees
- Support Vector Machine (SVM): A supervised learning model that analyzes data for classification
Both models are trained on the Cleveland Heart Disease Dataset from the UCI Machine Learning Repository.
- Data preprocessing and cleaning
- Model training with hyperparameter tuning using GridSearchCV
- Feature importance visualization
- Model evaluation with confusion matrices
- Web interface for making predictions
- RESTful API for integration with other systems
- Ensemble approach combining predictions from both models
The web application provides an easy-to-use interface where you can:
- Enter patient data
- Get predictions from both models
- View an ensemble prediction with risk assessment
- Explore feature importance and model performance
This project is licensed under the MIT License - see the LICENSE file for details.
- Cleveland Heart Disease Dataset: UCI Machine Learning Repository
- Scikit-learn for machine learning implementation
- Flask for web application framework