Repeats 14 yes-or-no questions about an auto-insurance claim and compares probability distributions and application decisions after uncertain values are sent to a human.
Original by TypeSafe AIEvaluationTypeSafe cookbookSource reviewed
Before you dive in
What you’ll find in the original
Keep the underlying Noul probabilities visible when routing to review.
Examine the automatic-action rate alongside consistency.
Use the cached samples to reproduce plots without mistaking a rerender for a new live experiment.
Worth knowing
Repeated agreement and human review bands are not substitutes for measuring real decision accuracy.