New belief function based methods for multi-criteria decision-making

Abstract

Any decision is closely linked to the quality and availability of information. Innovative methodologies dealing with imperfect information provided by more or less reliable and conflicting sources are proposed to help decision-makers. Our methods combine new uncertainty theories within Multi-Criteria Decision-Making methods. In our evidential reasoning based approach for multi-criteria decision analysis, we use the fuzzy sets, the possibility and belief function theories as conceptual analytical frameworks. In DSmT-AHP, we replace the initial AHP aggregation principles by a fusion process with the introduction of discounting factors able to discriminate importance and reliability of criteria. In our COWA-ER, we propose a cautious Ordered Weighted Averaging method for multi-criteria decision making under uncertainty. All these methods fit well in the domain of expert assessment for the context of natural hazards in mountains. We present a synthesis of the three aforementioned methods which are detailed in papers listed in the bibliography. © 2012 Springer-Verlag.