« Back to Glossary Index

A machine learning technique where models are trained using only a small number (k) of labeled examples per class, typically ranging from 1 to 5.