Reducing the overconfidence of base classifiers when combining their decisions

Download
  1. (PDF, 482 KB)
  2. Get@NRC: Reducing the overconfidence of base classifiers when combining their decisions (Opens in a new window)
DOIResolve DOI: http://doi.org/10.1007/3-540-44938-8_7
AuthorSearch for: ; Search for: ; Search for:
TypeBook Chapter
Proceedings titleMultiple Classifier Systems : 4th International Workshop, MCS 2003 Guildford, UK, June 11–13, 2003 Proceedings
Series titleLecture Notes In Computer Science; Volume 2709
Conference4th International Workshop on Multiple Classifier Systems (MCS 2003), June 11-13, 2003, Guildford, United Kingdom
ISSN0302-9743
ISBN978-3-540-40369-2
978-3-540-44938-6
Pages6573; # of pages: 9
Subjectmultiple classification systems; fusion rule; BKS method; local classifiers; sample size; apparent error; complexity; stacked generalization
AbstractWhen the sample size is small, the optimistically biased outputs produced by expert classifiers create serious problems for the combiner rule designer. To overcome these problems, we derive analytical expressions for bias reduction for situations when the standard Gaussian density-based quadratic classifiers serve as experts and the decisions of the base experts are aggregated by the behavior-space-knowledge (BKS) method. These reduction terms diminish the experts’ overconfidence and improve the multiple classification system’s generalization ability. The bias-reduction approach is compared with the standard BKS, majority voting and stacked generalization fusion rules on two real-life datasets for which the different base expert aggregates comprise the multiple classification system.
Publication date
PublisherSpringer Berlin Heidelberg
AffiliationNRC Institute for Biodiagnostics; National Research Council Canada
Peer reviewedYes
NRC number2055
NPARC number9148007
Export citationExport as RIS
Report a correctionReport a correction
Record identifier8a1144bd-69c8-49fe-9ae2-8ef86f388615
Record created2012-10-22
Record modified2016-08-02
Bookmark and share
  • Share this page with Facebook (Opens in a new window)
  • Share this page with Twitter (Opens in a new window)
  • Share this page with Google+ (Opens in a new window)
  • Share this page with Delicious (Opens in a new window)