Diabetes risk predictor

The brief

A health check-up centre wants to tell people early if they are at risk of diabetes. Build classification models on check-up measurements, compare them, and show which factors drive risk, so a doctor can explain the result to a patient. Good looks like: honest validation, sensible handling of missing values, a model a doctor can understand, and clear risk factors.

What you will build

1. Explore the data and fix impossible zero values
2. Train logistic regression, decision trees and gradient boosting
3. Evaluate with cross-validation, recall and AUC
4. Explain the risk factors
5. Build a simple risk calculator with a "see your doctor" message

Free, open data to build it with

• CDC diabetes health indicators ↗ UCI Machine Learning Repository

Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.