Healthcare Cost Prediction Engine
An interpretable model that quantifies the real drivers behind medical charges.
Overview
Medical charges feel like a black box to the people paying them. I built a model that not only predicts cost but explains it — because in healthcare, a number without a reason isn't trustworthy.
I engineered features from demographic and clinical factors, then trained interpretable regression models, prioritizing transparency over a marginal accuracy gain. Every prediction comes with the top drivers pushing it up or down.
Packaged behind a simple interface, this is the engine a quoting, budgeting, or planning tool would run on — giving payers and patients a defensible estimate instead of a surprise bill.
At a glance
Payers and patients struggle to anticipate medical charges, making budgeting and pricing opaque.
An interpretable regression model that quantifies how demographics and clinical factors drive cost.
A transparent cost estimator that surfaces the top drivers — directly usable in a quoting or planning tool.
Highlights
- Feature engineering across demographic and clinical signals
- Interpretable regression with per-prediction driver breakdowns
- Validated for transparency, not just headline accuracy
- Designed to plug into a quoting or planning product