Bias-variance tradeoff

The bias-variance tradeoff describes the tension between a model that is too simple and makes systematic errors, and one that is too sensitive to its training data and fails to generalize.

Bias-variance tradeoff sits in the Foundations part of the Agentik {OS} glossary, which defines the words used to build and run AI agent systems.

Also called Bias-variance dilemma.