Calibrating Margin-Based Classifier Scores into Polychotomous Probabilities

  • Martin Gebel
  • Claus Weihs
Conference paper
Part of the Studies in Classification, Data Analysis, and Knowledge Organization book series (STUDIES CLASS)

Abstract

Margin-based classifiers like the SVM and ANN have two drawbacks. They are only directly applicable for two-class problems and they only output scores which do not reflect the assessment uncertainty. K-class assessment probabilities are usually generated by using a reduction to binary tasks, univariate calibration and further application of the pairwise coupling algorithm. This paper presents an alternative to coupling with usage of the Dirichlet distribution.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Martin Gebel
    • 1
  • Claus Weihs
    • 2
  1. 1.Graduiertenkolleg Statistische Modellbildung, Lehrstuhl für Computergestützte StatistikUniversitÄt DortmundDortmundGermany
  2. 2.Lehrstuhl für Computergestützte StatistikUniversitÄt DortmundDortmundGermany

Personalised recommendations