Total: 1
In this paper, we consider the recently introduced application of the Deck of Cards Method (DCM) to ordinal regression proposing an extension to Robust Ordinal Regression and Stochastic Multiattribute Acceptability Analysis. In Multiple Criteria Decision Aiding context, the proposed methodology permits to assign a value to each alternative evaluated on a set of criteria hierarchically structured. The Decision Maker can provide precise or imprecise information at different levels of the hierarchy of criteria using the classical DCM framework. This information is therefore used to infer a value function compatible with it. The compatible value function can be a simple weighted sum, a piecewise linear value function, a general monotonic value function, or a Choquet integral. To provide robust recommendations to the Decision Maker, we consider the Robust Ordinal Regression and the Stochastic Multicriteria Acceptability Analysis because, even if in different ways, both of them take into account the whole set of models compatible with the preference information provided by the Decision Maker. The applicability of the proposed methodology is shown by a didactic example in which Italian regions are evaluated on criteria representing Circular Economy, Innovation Driven Development and Smart Specialization Strategies.