🧠 Insurance Charges Prediction App (Streamlit + XGBoost)

A fast and interactive web application built using Streamlit, this app predicts insurance charges based on user health and demographic inputs. It leverages an XGBoost model trained on real-world data and features an intuitive UI for quick experimentation.

🚀 Features

• Predict insurance charges based on:
  • Age, Gender, BMI
  • Smoking Status
  • Number of Children
  • Region
• Trained XGBoost model using insurance.csv dataset
• Preprocessing includes:
  • One-hot encoding for categorical fields
  • MinMaxScaler for numerical inputs
• Real-time prediction output with result messaging
• Multiple language support (English, 中文)

🧰 Tech Stack

• Frontend: Streamlit
• Model: XGBoost
• Preprocessing: scikit-learn
• Language: Python

📂 Folder Structure

insurance-app/
├── app.py                  # Main Streamlit app
├── model_training.py       # Model training script
├── xgb_model.pkl           # Trained XGBoost model
├── scaler.pkl              # Fitted MinMaxScaler
├── encoder.pkl             # Fitted OneHotEncoder
├── assets/
│   └── insurance-app-screenshot.png
├── requirements.txt
└── README.md

▶️ Running Locally

pip install -r requirements.txt
streamlit run app.py

Visit: http://localhost:8501

🌐 Deployment Options

You can deploy this app to:

• Streamlit Community Cloud
• Heroku
• Alibaba Cloud ECS
• Docker
• AWS EC2

📊 Sample Prediction Output

• Inputs: 35 yrs old, BMI: 24.5, Smoker: No, Region: southeast
• Predicted Insurance Charge: $7,852.34

🔒 Note

Ensure the following files are in the same directory as app.py when running:

• xgb_model.pkl
• scaler.pkl
• encoder.pkl

📬 Contact

Feel free to fork or contribute. For bugs or suggestions, please open an issue or submit a PR!