AI Glossary

Undersampling

What does “Undersampling” mean?

Undersampling: Reducing the number or frequency of examples from an overrepresented group.

Undersampling visual guide: Reducing the number or frequency of examples from an overrepresented group.
Visual guide Open full size (new tab)

What is an example of “Undersampling”?

A team samples only part of a very common class to build a more balanced training set.

Keep learning

See also

DatasetTermTraining dataTermData labelingTermClass imbalanceTermOversamplingTerm

Same topic

Hard negativeTermWeak supervisionTermValidation and test setsTermUnderfittingTermValue functionTerm