Integrating Dynamic Bayesian Networks and Analytic Hierarchy Process for Time-Dependent Multi-Criteria Decision-Making

Abstract

The analytic hierarchy process (AHP) has been a widely used method for handling multi-criteria decision-making (MCDM) problems since the 1980s. However, it postulates that criteria are independent and static, which may not always hold true in realistic situations. Although several methods have been proposed to relax the assumption of independence between criteria in the AHP, such as the analytic network process (ANP), these methods do not account for time-dependent criteria in the AHP. Consequently, this paper presents an innovative method that integrates dynamic Bayesian networks (DBNs) with the AHP to model dynamic interdependencies between criteria in MCDM problems. We illustrate the proposed method through a comprehensive numerical example and compare the result with the conventional AHP. The findings suggest that the proposed method extends the AHP to accommodate time-dependent issues and, when ignoring specific information, reduces to the conventional AHP, thereby demonstrating that our approach serves as a more general AHP model.