JPL Technical Report Server

Active learning in the presence of unlabelable examples

Show simple item record

dc.contributor.author Mazzoni, Dominic
dc.contributor.author Wagstaff, Kiri
dc.date.accessioned 2006-07-10T15:47:09Z
dc.date.available 2006-07-10T15:47:09Z
dc.date.issued 2004-09-20
dc.identifier.citation European Conference on Machine Learning, Pisa, Italy, September 20, 2004. en
dc.identifier.clearanceno 04-1403
dc.identifier.uri http://hdl.handle.net/2014/39483
dc.description.abstract We propose a new active learning framework where the expert labeler is allowed to decline to label any example. This may be necessary because the true label is unknown or because the example belongs to a class that is not part of the real training problem. We show that within this framework, popular active learning algorithms (such as Simple) may perform worse than random selection because they make so many queries to the unlabelable class. We present a method by which any active learning algorithm can be modified to avoid unlabelable examples by training a second classifier to distinguish between the labelable and unlabelable classes. We also demonstrate the effectiveness of the method on two benchmark data sets and a real-world problem. en
dc.description.sponsorship NASA/JPL en
dc.format.extent 1157736 bytes
dc.format.mimetype application/pdf
dc.language.iso en_US en
dc.publisher Pasadena, CA : Jet Propulsion Laboratory, National Aeronautics and Space Administration, 2004 en
dc.subject machine learning en
dc.subject active learning en
dc.title Active learning in the presence of unlabelable examples en
dc.type Preprint en


Files in this item

This item appears in the following Collection(s)

Show simple item record

Search


Browse

My Account