An evolutionary approach for inferring the model parameters of the hierarchical ELECTRE III method

Given a finite set of alternatives, the ranking problem statement builds a preference preorder (partial or complete) on this set. In this paper, we are interested in multiple criteria ranking problems with a hierarchical structure of criteria; more precisely, we are interested in the existing hierarchical ELECTRE III method. This method requires eliciting several preference parameters (namely, the weights and the veto thresholds). A direct elicitation of such parameters can be cognitively very demanding; thus, it is adequate to define the parameters in a way that requires much less cognitive effort from the decision-maker. The model parameters can be indirectly elicited by using holistic information provided by the decision-maker; this information can be given in the form of a ranking on a set of reference alternatives and some additional preference information. This paper proposes an aggregation-disaggregation approach for inferring the model parameters of the hierarchical ELECTRE III based on an evolutionary algorithm. To verify the applicability and validity of the proposed preference disaggregation methodology, an illustrative example is addressed regarding the ranking of a set of universities. (c) 2022 Elsevier Inc. All rights reserved.

Automated summary

This paper focuses on hierarchical multiple criteria ranking problems and proposes an aggregation-disaggregation approach using an evolutionary algorithm to infer the model parameters. The proposed method allows indirect elicitation of the parameters through a ranking of reference alternatives and additional preference information.