This is a multimodel for predicting heart disease. The multimodel is a combination of 3 models: Logistic Regression, Random Forest, and XGBoost. The multimodel is trained on the BKDS Heart Disease Dataset. The dataset contains 76 attributes, but all published experiments refer to using a subset of 14 of them. The "goal" field refers to the presence of heart disease in the patient. It is integer valued from 0 (no presence) to 4. Experiments with the Cleveland database have concentrated on simply attempting to distinguish presence (values 1,2,3,4) from absence (value 0).
To install the multimodel, run the following command:
git clone bkds04
cd bkds04
pip install -r requirements.txt
To run the multimodel, run the following command:
streamlit run apps/app.py