How shreyask5/stroke-risk-prediction works
Project type and purpose
• This is a full-stack web application called “Stroke-Risk-Prediction.”
• The frontend is a single-page application (SPA) built with Vite + React + TypeScript and a custom UI component library.
• The backend is a Python service (Flask or similar) running under Gunicorn, which loads a pre-trained ML model (joblib) to predict stroke risk based on user inputs.
File structure
• Top-level:
• docker-compose.yml orchestrates two services: frontend and backend.
• backend/ contains the prediction service, ML assets, and training code.
• frontend/ contains all client-side code, configs, and UI components.
• backend/:
• app.py – the API server entrypoint.
• gunicorn.conf.py – Gunicorn settings.
• requirements.txt – Python dependencies.
• stroke_prediction_model.joblib – serialized ML model.
• stroke-data.csv & training.py – data and training pipeline.
• frontend/:
• src/main.tsx, src/App.tsx – SPA bootstrap and root component.
• src/components/StrokePredictionForm.tsx – the form for user input.
• src/components/ui/… – reusable UI primitives (buttons, dialogs, inputs, etc.).
• package.json & bun.lockb – JS/TS dependencies.
• vite.config.ts, tsconfig*.json – build & TS configuration.
README
• Minimal README. The project name indicates its function: “Stroke-Risk-Prediction.”