Towards robust rank correlation measures for numerical observations on the basis of fuzzy orderings

Ulrich Bodenhofer, Frank Klawonn

Research output: Chapter in Book/Report/Conference proceedingsConference contributionpeer-review

4 Citations (Scopus)

Abstract

This paper aims to demonstrate that established rank correlation measures are not ideally suited for measuring rank correlation for numerical data that are perturbed by noise. We propose a robust rank correlation measure on the basis of fuzzy orderings. The superiority of the new measure is demonstrated by means of illustrative examples.

Original languageEnglish
Title of host publicationNew Dimensions in Fuzzy Logic and Related Technologies - Proceedings of the 5th EUSFLAT 2005 Conference
PublisherVSB-Technical University of Ostrava
Pages321-327
Number of pages7
ISBN (Print)9788073683863
Publication statusPublished - 2007
Externally publishedYes
Event5th Conference of the European Society for Fuzzy Logic and Technology, EUSFLAT 2007 - Ostrava, Czech Republic
Duration: 11 Sept 200714 Sept 2007

Publication series

NameNew Dimensions in Fuzzy Logic and Related Technologies - Proceedings of the 5th EUSFLAT 2005 Conference
Volume1

Conference

Conference5th Conference of the European Society for Fuzzy Logic and Technology, EUSFLAT 2007
Country/TerritoryCzech Republic
CityOstrava
Period11.09.200714.09.2007

Keywords

  • Fuzzy Orderings
  • Rank Correlation
  • Robust Statistics

Fingerprint

Dive into the research topics of 'Towards robust rank correlation measures for numerical observations on the basis of fuzzy orderings'. Together they form a unique fingerprint.

Cite this