Proportional k-interval discretization for naive-Bayes classifiers

Yang, Ying and Webb, Geoffrey I. 2001, Proportional k-interval discretization for naive-Bayes classifiers, in ECML 2001 : Machine Learning : 12th European Conference on Machine Learning, Springer-Verlag, Berlin, Germany, pp. 564-575, doi: 10.1007/3-540-44795-4.

Attached Files
Name Description MIMEType Size Downloads

Title Proportional k-interval discretization for naive-Bayes classifiers
Author(s) Yang, Ying
Webb, Geoffrey I.
Conference name European Conference on Machine Learning (12th : 2001 : Freiburg, Germany)
Conference location Freiburg, Germany
Conference dates 3-7 September 2001
Title of proceedings ECML 2001 : Machine Learning : 12th European Conference on Machine Learning
Editor(s) Carbonell, Jaime G.
Siekmann, Jorg
Publication date 2001
Series Lecture notes in computer science ; 2167
Start page 564
End page 575
Publisher Springer-Verlag
Place of publication Berlin, Germany
Summary This paper argues that two commonly-used discretization approaches, fixed k-interval discretization and entropy-based discretization have sub-optimal characteristics for naive-Bayes classification. This analysis leads to a new discretization method, Proportional k-Interval Discretization (PKID), which adjusts the number and size of discretized intervals to the number of training instances, thus seeks an appropriate trade-off between the bias and variance of the probability estimation for naive-Bayes classifiers. We justify PKID in theory, as well as test it on a wide cross-section of datasets. Our experimental results suggest that in comparison to its alternatives, PKID provides naive-Bayes classifiers competitive classification performance for smaller datasets and better classification performance for larger datasets.
ISBN 3540425365
Language eng
DOI 10.1007/3-540-44795-4
Field of Research 080199 Artificial Intelligence and Image Processing not elsewhere classified
Socio Economic Objective 970109 Expanding Knowledge in Engineering
HERDC Research category E1 Full written paper - refereed
Copyright notice ┬ęSpringer-Verlag Berlin Heidelberg 2001
Persistent URL

Connect to link resolver
Unless expressly stated otherwise, the copyright for items in DRO is owned by the author, with all rights reserved.

Version Filter Type
Citation counts: TR Web of Science Citation Count  Cited 0 times in TR Web of Science
Scopus Citation Count Cited 0 times in Scopus
Google Scholar Search Google Scholar
Access Statistics: 469 Abstract Views, 0 File Downloads  -  Detailed Statistics
Created: Mon, 07 Jul 2008, 09:37:14 EST

Every reasonable effort has been made to ensure that permission has been obtained for items included in DRO. If you believe that your rights have been infringed by this repository, please contact