Robust Ordinal Regression in case of Imprecise Evaluations

TL;DR

This paper extends Robust Ordinal Regression in MCDA to handle imprecise evaluations, allowing for more comprehensive preference relations that consider uncertainty in alternative assessments.

Contribution

It introduces a novel extension of ROR that incorporates imprecise evaluations, broadening the applicability of MCDA methods under uncertainty.

Findings

Enhanced preference relations accounting for evaluation imprecision

Framework for necessary and possible preferences with imprecise data

Improved decision support in uncertain evaluation scenarios

Abstract

Robust Ordinal Regression (ROR) is a way of dealing with Multiple Criteria Decision Aiding (MCDA), by considering all sets of parameters of an assumed preference model, that are compatible with preference information given by the Decision Maker (DM). As a result of ROR, one gets necessary and possible preference relations in the set of alternatives, which hold for all compatible sets of parameters or for at least one compatible set of parameters, respectively. In this paper, we extend the MCDA methods based on ROR, by considering one important aspect of decision problems: imprecise evaluations. To deal with imprecise evaluations of some alternatives on particular criteria, we extend the set of considered variables to define necessary and possible preference relations taking into account this imprecision. In consequence, the concepts of necessary and possible preference represent not only all compatible sets of parameters, but also all possible values of the imprecise evaluations of alternatives.