Evaluation of an Interval-Type Model on Fuzzy Regression

A fuzzy regression model is classified into two types: non-interval-type and interval-type fuzzy regressions. A non-interval-type fuzzy regression model can analyze errors similar to the manner in which a statistical least squares model can. In contrast, because an interval-type fuzzy regression model illustrates the possibility of an analyzed system by including data, the obtained regression does not analyze prediction accuracies such as in error analysis. In other words, it is important to illustrate the amount of possibilities of an analyzed system by regression outputs. The appropriate evaluation functions, which can be easily interpreted, are used for this purpose. This paper proposes a new evaluation function, which is validated using a numerical example. The evaluation function is explained and discussed herein using the numerical example.