AI Model Backend

A FastAPI-based REST API for disease prediction using a machine learning model with 24 input features.

Project Structure

backend/
├── app/
│   ├── __init__.py
│   ├── predict.py          # Model loading and prediction logic
│   └── schemas.py          # Pydantic models for request/response validation
├── model/
│   └── disease prediction model.pkl  # Trained ML model
├── main.py                 # FastAPI application entry point
├── requirements.txt        # Python dependencies
└── README.md

Features

• FastAPI-based REST API
• CORS support for cross-origin requests
• Machine learning model for disease prediction
• 24-feature input validation
• Confidence score for predictions
• Hot reload support for development

API Endpoints

GET /

Returns API information and usage instructions.

POST /predict

Accepts 24 float features and returns a prediction with confidence score.

Request body:

{
  "features": [float1, float2, ..., float24]
}

Response:

{
  "prediction": "predicted_class",
  "confidence": 0.95
}

Requirements

• Python 3.11 or higher
• pip (Python package installer)
• Git

Installation and Setup

Clone the Repository

git clone https://github.com/HackSprint2025/ai-model-backend.git
cd ai-model-backend

Create Virtual Environment

python -m venv venv

Activate Virtual Environment

On Windows:

venv\Scripts\activate

On macOS/Linux:

source venv/bin/activate

Install Dependencies

pip install -r requirements.txt

Running the Application

Start the Development Server

uvicorn main:app --reload --host 0.0.0.0 --port 8000

The API will be available at:

• Local: http://localhost:8000
• Network: http://0.0.0.0:8000

Access API Documentation

Once the server is running, access the interactive API documentation at:

• Swagger UI: http://localhost:8000/docs
• ReDoc: http://localhost:8000/redoc

Technology Stack

• FastAPI: Web framework for building APIs
• Uvicorn: ASGI server
• Pydantic: Data validation using Python type annotations
• NumPy: Numerical computing
• Scikit-learn: Machine learning library
• Pickle: Model serialization

Development

The application uses FastAPI's hot reload feature. Any changes to the Python files will automatically restart the server during development.

CORS Configuration

The API is configured to accept requests from all origins. For production deployment, update the FRONTEND_ORIGINS variable in main.py to specify allowed origins.