Resource title

Leave one out error, stability, and generalization of voting combinations of classifiers

Resource image

image for OpenScout resource :: Leave one out error, stability, and generalization of voting combinations of classifiers

Resource description

The authors study the generalization error of voting combinations of learning machines. A special case considered is bagging. They analyze in detail combinations of kernel machines, such as support vector machines, and present theoretical bounds on their generalization error using leave one out error estimates. They also derive novel bounds on the stability of combinations of any classifiers. These bounds can be used to formally show that for example, bagging increases the stability of unstable learning machines. As a special case they study the stability and generalization of bagging kernel machines and report experiments validating the theoretical findings.

Resource author

Resource publisher

Resource publish date

Resource language

en

Resource content type

application/pdf

Resource resource URL

http://flora.insead.edu/fichiersti_wp/inseadwp2001/2001-21.pdf

Resource license

Copyright INSEAD. All rights reserved