A general approach to fuzzy regression models based on different loss functions

Abstract

In this paper, a new general approach is presented to fit a fuzzy regression model when the response variable and the parameters of model are as fuzzy numbers. In this approach, for estimating the parameters of fuzzy regression model, a new definition of objective function is introduced based on the different loss functions and under the averages of differences between the $$\alpha $$ -cuts of errors. The application of the proposed approach is studied using a simulated data set and some real data sets in the presence of different types of outliers.