🫀 Cardio Risk ML System

A full-stack machine learning web application that predicts the risk of cardiovascular disease based on patient health metrics.
The system combines custom ML model training, Flask backend APIs, JWT authentication, and a React (Vite) frontend dashboard, all deployed on free cloud platforms.

---

🚀 Live Demo

• Frontend (Vercel):
  👉 https://your-frontend-url.vercel.app

• Backend API (Render):
  👉 https://cardio-risk-ml-system.onrender.com

---

📌 Features

🔐 Authentication

• User registration & login
• JWT-based secure authentication
• Persistent user sessions

🧠 Machine Learning

• Custom Decision Tree implementation
• Feature engineering (BMI, pulse pressure, health index)
• Train–Test Split + Stratified K-Fold Cross Validation
• Best model selection based on accuracy
• Model saved and reused for real-time prediction

📊 Dashboard

• Individual user prediction history
• Risk score trend visualization
• Latest health stats summary
• Daily health tips

⚙️ Backend APIs

• /api/auth/login
• /api/auth/register
• /api/predict
• /api/user/history
• /api/user/stats

---

🏗️ Tech Stack

Frontend

• React (Vite)
• Tailwind CSS
• React Router
• Fetch API

Backend

• Python
• Flask
• Flask SQLAlchemy
• JWT Authentication
• SQLite

Machine Learning

• NumPy
• Pandas
• scikit-learn (CV & data splitting)
• Custom Decision Tree (no sklearn model)

Deployment

• Frontend: Vercel
• Backend: Render
• Version Control: Git & GitHub

---