Issue: 2010/Vol.20/No.2, Pages 25-40

A METHOD FOR DETECTING OUTLIERS IN FUZZY REGRESSION

Barbara Gładysz

Full paper (PDF)    RePEC

Cite as: B. Gładysz. A method for detecting outliers in fuzzy regression. Operations Research and Decisions 2010: 20(2), 25-40.

Abstract
In this article we propose a method for identifying outliers in fuzzy regression. Outliers in a sample may have an important influence on the form of the regression equation. For this reason there is great scientific interest in this issue. The method presented is analogous to the method of finding outliers based on the studentized distribution of residuals. In order to identify outliers, regression models are constructed with an additional explanatory variable for each observation. Next, the significance of a fuzzy regression coefficient is analysed considering this additional explanatory variable. Illustrative examples are presented.

Keywords: fuzzy regression, outliers, possibility theory

Received:     Accepted: