Outlier detection approaches in fuzzy regression models

In this paper, we propose three outlier detection approaches for the fuzzy regression models proposed by Tanaka after a brief review of the related literatures. Generally speaking, for the upper regression model, the aim is to pick out some abnormal data that is not consistent with the trend of the upper regression model; for the lower regression model, as it often has no feasible solutions, the efforts are made to identify the data that has effect on the infeasibility of the lower regression model.