Building a Type-2 Fuzzy Regression Model based on Creditability Theory

Information in real life may have linguistically vagueness. Thus, type-1 fuzzy set was introduced to model this uncertainty. Additionally, same words will mean variously to different people, which means uncertainty also exists when associated with the membership function of a type-1 fuzzy set. Type-2 fuzzy set is then invented to express the hybrid uncertainty of both primary fuzziness and secondary one of membership functions. On the one hand, type-2 fuzzy variable models the vagueness of information better. On the other hand, those variables are hard to deal with its three-dimensional feature given. To address problems in presence of such variables with hybrid fuzziness, a new class of type-2 fuzzy regression model is built based on credibility theory, and is called the T2 fuzzy expected value regression model. The new model will be developed into two forms: form-A and form-B. This paper is a further work based on our former research of type-2 fuzzy qualitative regression model.