The same 14 parameters are almost always used to detect heart disease. This project aims to discover whether any of the 62 other possible parameters have an impact on heart disease. My team and I developed a logistic regression model that was able to detect the most significant indicators of cardiovascular disease out of the set of 76 health-related factors using Pandas and Scikit-Learn during the 24-hour Daisy Intelligence Hackathon at the University of Toronto. Our results suggest that the 62 unused parameters are very significant because they are key in determining the severity of heart disease.
Key Documents:
- Daisy-Heart-Disease.ipynb
- Daisy Intelligence Elevator Pitch - Aha!AI.pdf