Classical ML

Expected Calibration Error (ECE)

Binned average of |accuracy − confidence|, weighted by bin size.

Why interviewers ask about it

Calibration drifts before discrimination does, which makes ECE an excellent early drift alarm in production.

The property that predicted probabilities match observed frequencies - among all cases scored 0.7, about 70% are positive.

This term is part of the free AI/ML Engineer interview preparation module - browse the full glossary for every definition.