- Project
- AI/ML
- Web
- Team
- 2024
Medicine recommendation from symptoms
Type your symptoms and get a likely disease with its description, precautions, medications, diet, and workout. A Flask app on top of five compared classifiers.
How it works
- 014,920 records: 132 symptoms mapped to 41 diseases
- 02SVC, Random Forest, Gradient Boosting, KNN, and Naive Bayes compared; linear SVC served
- 03Lookup tables add description, precautions, medications, diet, and workout per disease
- 04Flask app: typed symptoms in, six result panels out
The problem
Searching symptoms online returns a wall of possibilities. This project turns a symptom list into one likely disease and practical next steps.
My part
Trained and compared the models, built the Flask app, and wrote the 18-page report.
A course project for a team of three; I built the models, the Flask app, and the report myself.
Limitations
- A learning project on a public dataset. It is not medical advice.
