EPSILON-FOCUSING - A STRATEGY FOR ACTIVE EXAMPLE SELECTION

In most classical formulations of learning from examples, a passive le arner is presented with examples randomly drawn. Here we discuss an ep silon-focusing strategy that actively chooses examples for concept lea rning. We describe the local focused property that functions must have for such a strategy to work. Finally, we develop an example selection heuristic, motivated by epsilon-focusing, for training a computer vis ion system to perform a human face detection task. (C) 1998 Elsevier S cience B.V. All rights reserved.