📊 Insurance Data Analysis & Model Comparison

A Streamlit web application for analyzing insurance data and comparing different machine learning models for predicting insurance charges.

🌟 Features

• Data Upload: Easily upload your own CSV dataset for analysis
• Data Visualization: Generate various plots to understand your data
  • Distribution plots
  • Scatter plots with optional color variables
  • Box plots for categorical analysis
  • Violin plots with data points
  • Correlation heatmaps for numeric features
• Model Training: Compare different regression models
  • Linear Regression
  • Decision Tree
  • Random Forest
• Model Evaluation: View performance metrics
  • Root Mean Squared Error (RMSE)
  • Mean Absolute Error (MAE)
  • R-squared (R²)
• Custom Predictions: Test trained models with your own input values
  • Support for both numeric and categorical inputs
  • Includes demographic features (sex, smoker status, region)

📋 Requirements

• Python 3.8+
• Streamlit
• Pandas
• NumPy
• Matplotlib
• Seaborn
• Scikit-learn
• Plotly Express

🚀 Installation

1. Clone this repository:

   git clone https://github.com/adi99304/Insurance_prediction.git
   cd Insurance_prediction

2. Create a virtual environment (optional but recommended):

   python -m venv venv
   
   # On Windows
   venv\Scripts\activate
   
   # On macOS/Linux
   source venv/bin/activate

3. Install the required packages:

   pip install -r requirements.txt

4. Run the Streamlit app:

   streamlit run app.py

📝 Usage

1. Upload your CSV file containing insurance data

   • Required columns: age, sex, bmi, children, smoker, region, charges
   • You can use the sample dataset provided in the data folder

2. Select plot types and variables to visualize your data

3. Choose a model type and train it on your data

4. Enter custom values to test predictions with your trained model

📚 Example Dataset

The app works with insurance datasets containing the following features:

• age: Age of the insured person
• sex: Gender of the insured (male/female)
• bmi: Body Mass Index
• children: Number of dependents
• smoker: Smoking status (yes/no)
• region: Residential area (northeast/northwest/southeast/southwest)
• charges: Insurance premium charges

🔄 Converting from R Shiny

This app was converted from an R Shiny dashboard to Python Streamlit. The conversion maintains the core functionality while leveraging Python's machine learning ecosystem.

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

👨‍💻 Author

Your Name - (https://github.com/adi99304)

🙏 Acknowledgments

• Streamlit for the amazing framework
• Scikit-learn for machine learning tools
• Plotly Express for interactive visualizations