The problem
A screening tool has to be explicit about what it is not. It surfaces an indication to discuss with a clinician, not a diagnosis.
What I built
- Trained a random forest classifier and served it behind a FastAPI backend with a React client.
Stack
ReactViteFastAPIPythonRandom Forestscikit-learn
Skills applied
- Machine learning
- FastAPI
- React