Reducing the overconfidence of base classifiers when combining their decisions

DOI10.1007/3-540-44938-8_7MaRDI QIDQ4448266zbMATH OpenFDO

Authors Ray Somorjai, Richard Baumgartner, Šarūnas Raudys

Publication date 17 February 2004

Published in Lecture Notes in Computer Science (Search for Journal in Brave)

Full work available at URL http://link.springer.de/link/service/series/0558/bibs/2709/27090065.htm, http://nrc-publications.canada.ca/eng/view/object/?id=8a1144bd-69c8-49fe-9ae2-8ef86f388615

Mathematics Subject Classification ID

Pattern recognition, speech recognition (68T10) Computing methodologies and applications (68U99)

Recommendations

• Reduction of the boasting bias of linear experts
• Trainable fusion rules. II. Small sample-size effects
• Linear and order statistics combiners for pattern classification
• Confidence combination methods in multi-expert systems
• The behavior knowledge space fusion method: Analysis of generalization error and strategies for performance improvement

Cited in

(2)

• Reduction of the boasting bias of linear experts
• The behavior knowledge space fusion method: Analysis of generalization error and strategies for performance improvement

This page was built for publication: Reducing the overconfidence of base classifiers when combining their decisions

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4448266)