Ranking methods based on dominance measures accounting for imprecision

Abstract

The additive multi-attribute utility model is widely used in multicriteria decision-making. However, it is often not easy to elicit precise values for the scaling weights representing the relative importance of criteria. In a group decision-making context a very widespread approach is to derive incomplete information, such as weight intervals or ordinal information rather than precise weights from a negotiation process. Different approaches have been proposed to deal with such situations. We advance two approaches based on dominance measures accounting for imprecise weights and compare them with other existing approaches using Monte Carlo simulation. © 2009 Springer-Verlag Berlin Heidelberg.