Committee-Based Selective Sampling with Parameters Set by Meta-Learning

Warning

This publication doesn't include Faculty of Sports Studies. It includes Faculty of Informatics. Official publication website can be found on muni.cz.

Authors	NEPIL Miloslav POPELÍNSKÝ Lubomír
Year of publication	2002
Type	Article in Proceedings
Conference	Proceedings of the second IDDM workshop at ECML/PKDD 2002
MU Faculty or unit	Faculty of Informatics
Citation
Web	IDDM 2002 - online proceedings
Field	Use of computers, robotics and its application
Keywords	selective sampling; meta-learning
Description	This contribution presents a parametric variant of committee-based selective sampling. The committee members learned on small subsets obtained by random sampling from the original dataset are used to classify the rest of the dataset. In the second part of the paper a meta-learning technique for parameter setting is introduced and experimental results obtained with it are discussed.
Related projects:	Human-computer interaction, dialog systems and assistive technologies