EVALUATION OF FUZZY LINEAR-REGRESSION MODELS BY COMPARING MEMBERSHIP FUNCTIONS

Authors
KIM BJ BISHU RR
Citation
Bj. Kim et Rr. Bishu, EVALUATION OF FUZZY LINEAR-REGRESSION MODELS BY COMPARING MEMBERSHIP FUNCTIONS, Fuzzy sets and systems, 100(1-3), 1998, pp. 343-352
Citations number
13
Categorie Soggetti
Statistic & Probability",Mathematics,"Computer Science Theory & Methods","Statistic & Probability",Mathematics,"Computer Science Theory & Methods
Journal title
Fuzzy sets and systems → ACNP
ISSN journal
01650114
Volume
100
Issue
1-3
Year of publication
1998
Pages
343 - 352
Database
ISI
SICI code
0165-0114(1998)100:1-3<343:EOFLMB>2.0.ZU;2-0
Abstract
Fuzzy linear regression models can provide an estimated fuzzy number t hat has a fuzzy membership function. If a point that has the highest m embership value from the estimated fuzzy number is not within the supp ort of the observed fuzzy membership function, a decision-maker can ha ve high risk from the estimate. In this study a modification of fuzzy linear regression analysis based on a criterion of minimizing the diff erence of the fuzzy membership values between the observed and estimat ed fuzzy numbers is proposed. Two numerical examples are used to evalu ate the fuzzy regression models. (C) 1998 Elsevier Science B.V. All ri ghts reserved.