Cohen’s Kappa

A metric measuring agreement between raters/models accounting for chance agreement.

This concept is essential for understanding model evaluation & metrics and forms a key part of modern AI systems.