Ridge regression; Multicollinearity; Ridge fuzzy regression model; Fuzzy multiple linear regression model
Issue Date
2019-10
Publisher
Springer Berlin Heidelberg
Citation
International Journal of Fuzzy Systems, v. 21, No. 7, Page. 2077–2090
Abstract
Ridge regression model is a widely used model with many successful applications, especially in managing correlated covariates in a multiple regression model. Multicollinearity represents a serious threat in fuzzy regression models as well. We address this issue by combining ridge regression with the fuzzy regression model. Our proposed algorithm uses the α-level estimation method to evaluate the parameters of the ridge fuzzy regression model. Two examples are given to illustrate the ridge fuzzy regression model with crisp input/fuzzy output and fuzzy coefficients.