Health Insurance Analysis

Notebook-based Python analysis of U.S. medical insurance costs and the variables that most strongly influence insurance charges.

Project focus

This project explores how insurance charges shift across factors such as:

• age
• BMI
• smoking status
• number of children
• region

The notebook combines exploratory analysis with predictive modeling so the work does not stop at charts alone.

Methods used

• Exploratory data analysis with pandas, Seaborn, and Matplotlib
• Feature-level comparison of charge drivers, including smoker and BMI effects
• Predictive modeling with:
  • Linear Regression
  • Random Forest Regressor
  • Gradient Boosting Regressor
• Hyperparameter tuning with GridSearchCV for the gradient boosting model

Repo layout

• US Medical Insurance Costs Analysis.ipynb - the main analysis notebook

Run locally

Place the expected insurance.csv dataset next to the notebook, then run:

jupyter notebook "US Medical Insurance Costs Analysis.ipynb"

Why this repo is in the portfolio

This project shows a practical analysis flow: understand the dataset, identify meaningful business drivers, compare baseline and ensemble models, and use the notebook as both a working document and a communication artifact.