"On-line Active Learning with Enhanced Reliability Concepts"
: Proceedings of the IEEE Conference on Evolving and Adaptive Intelligent Systems, IEEE press, Seite(n) 1-6, 2012
On-line Active Learning with Enhanced Reliability Concepts
Sprache des Titels:
Proceedings of the IEEE Conference on Evolving and Adaptive Intelligent Systems
In this paper, we present a new methodology for conducting active learning in a single-pass on-line learning context, thus reducing the annotation effort for operators by selecting the most informative samples, Le. those ones helping incremental, evolving classifiers most to improve their own predictive performance. Our approach will be based on certtlintybased sample selection in connection with version-space reduction approach. Therefore, two new concepts regarding classifier's
reliability in its predictions will be investigated and developed in
connection with evolving fuzzy classifiers: co1fflict and ignorance.
Conflict models the extent to which a new query point Des in
the conflicting region between two or more classes. Ignorance
represents the extent to which the new query point appears in
an nnexplored region of the feature space. The results based
on real-world streaming classification data will show a stable
high predictive quaUty of our approach, despite the fact that the
requested number of class labels is decreased by up to 90%.