Projects

Projects

1. Medical Cost Prediction — Insurance Analytics

Built in R, this project develops and evaluates log-linear regression and gradient boosting models to predict annual medical expenditures using demographic and lifestyle factors. The work emphasizes model interpretability, statistical inference, and comparative performance evaluation across classical and machine learning approaches.

Tools: R, tidyverse, lm, sandwich, lmtest, gbm
Domain: Health Economics | Predictive Modeling | Insurance Analytics

2. Medical Cost Estimator Web App — Model Deployment

Built with Shiny and deployed via shinylive on GitHub Pages, this interactive web application estimates annual and monthly medical costs based on user health characteristics. The application operationalizes a log-linear regression model for real-time, transparent decision support.

Tools: R, Shiny, shinylive, GitHub Pages, AI-assisted development (ChatGPT)
Domain: Applied Machine Learning | Model Deployment | Health Data Science