Glossary term
A model's ability to rank cases with different outcome risk, distinct from probability calibration.
← Back to the glossary