AI Glossary

Weak supervision

What does “Weak supervision” mean?

Weak supervision: Creating training labels from imperfect rules, heuristics, or external sources.

Weak supervision visual guide: Creating training labels from imperfect rules, heuristics, or external sources.
Visual guide Open full size (new tab)

What is an example of “Weak supervision”?

A team combines several noisy labeling rules before training a classifier.

Keep learning

See also

DatasetTermTraining dataTermData labelingTermUndersamplingTermHard negativeTerm

Same topic

Active learningTermFederated learningTermWeight decayTermValue functionTermWeight initializationTerm