Adapting to drift in continuous domains (Extended abstract)

Miroslav Kubat, Gerhard Widmer

Research output: Chapter in Book/Report/Conference proceedingConference contribution

35 Scopus citations

Abstract

The experiments demonstrate that FRANN compares favourably with FLORA4 in the presence of concept drift. Learning is possible from examples described by symbolic as well as by numeric attributes, and because of its representation formalism (RBF networks, which realize a kind of prototype weighting scheme) FRANN is particularly effective in capturing concepts with nonlinear boundaries.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages307-310
Number of pages4
Volume912
ISBN (Print)3540592865, 9783540592860
StatePublished - 1995
Externally publishedYes
Event8th European Conference on Machine Learning, ECML 1995 - Heraclion, Greece
Duration: Apr 25 1995Apr 27 1995

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume912
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other8th European Conference on Machine Learning, ECML 1995
CountryGreece
CityHeraclion
Period4/25/954/27/95

    Fingerprint

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Kubat, M., & Widmer, G. (1995). Adapting to drift in continuous domains (Extended abstract). In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 912, pp. 307-310). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 912). Springer Verlag.