Active Learning on Data Streams

Instance Selection

Discriminating mechanism identifying the most informative samples from the stream to maximize model improvement per unit of labeling.

← Back