Active Learning Algorithms

When labels arrive in large streams, the learner has to update as they come instead of waiting for a full dataset. The page starts, and for now ends, with the Passive Aggressive classifier for massive data streams such as Twitter-scale text.

The same notes are in Active Learning and Online Learning.

Passive Aggressive classifier

For a stream that never stops, the first algorithm to reach for is a cheap one. The Passive Aggressive (PA) algorithm is perfect for classifying massive streams of data (e.g. Twitter). It is easy to implement and very fast, but does not provide global guarantees like the support-vector machine (SVM). To see it worked on text, Victor Lavrenko's Text Classification 3: Passive Aggressive Algorithm is the YouTube, seems like active learning in stream..? lecture.