www.uhasselt.be
DSpace

Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/16175

Title: Fuzzy data envelopment analysis in composite indicator contruction
Authors: Shen, Yongjun
Hermans, Elke
Brijs, Tom
Wets, Geert
Issue Date: 2014
Publisher: Springer Berlin Heidelberg
Citation: Emrouznejac, Ali; Tavana, Madjid (Ed.). Performance measurement with Fuzzy Data Envelopment analysis, p. 89-100
Series/Report: Studies in Fuzziness and Soft Computing
Series/Report no.: 309
Abstract: Data envelopment analysis (DEA) as a performance evaluation methodology has lately received considerable attention in the construction of composite indicators (CIs) due to its prominent advantages over other traditional methods. In this chapter, we present the extension of the basic DEA-based CI model by incorporating fuzzy ranking approach for modeling qualitative data. By interpreting the qualitative indicator data as fuzzy numerical values, a fuzzy DEA-based CI model is developed, and it is applied to construct a composite alcohol performance indicator for road safety evaluation of a set of European countries. Comparisons of the results with the ones from the imprecise DEA-based CI model show the effectiveness of the proposed model in capturing the uncertainties associated with human thinking, and further imply the reliability of using this approach for modeling both quantitative and qualitative data in the context of CI construction.
Notes: [Shen, Yongjun; Hermans, Elke; Brijs, Tom; Wets, Geert] Hasselt Univ, Transportat Res Inst IMOB, B-3590 Diepenbeek, Belgium.
URI: http://hdl.handle.net/1942/16175
DOI: 10.1007/978-3-642-41372-8_4
ISI #: 000343011500005
ISBN: 9783642413711
ISSN: 1434-9922
Category: B2
Type: Book Section
Validation: vabb, 2016
Appears in Collections: Research publications

Files in This Item:

Description SizeFormat
N/A313.75 kBAdobe PDF

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.