SW StudyWalks

Artificial Intelligence  /  AI 0433  ·  Atom · ~20 seconds

The Nudging Rule

Video not yet published
to the StudyWalks catalog
State

The perceptron learns by nudging: a correct verdict changes nothing, and every error shifts each weight a step toward what would have made the right answer.

Show

Called ripe when truly unripe: the score ran high, so subtract each feature's value from its weight, and one from the balance.

Watch for

Successes teach nothing — the rule is error-driven, local, and cheap by design.