An Extension of the Analytic Hierarchy Process Method under the Belief Function Framework

In this paper, an extension of the belief Analytic Hierarchy Process (AHP) method is proposed, based on the belief function framework. It takes into account the fact that the pair-wise comparison between criteria and alternatives may be uncertain and imprecise. Therefore, it introduces a new way to cope with expert judgments. Thus to express his preferences, the decision maker is allowed to use a belief assessment instead of exact ratios. The proposed extension also models the relationship between the alternative and criterion levels through conditional beliefs. Numerical examples explain in detail and illustrate the proposed approach.