2208. A study on first ranked alternatives in Promethee methods
Invited abstract in session MD-8: Methodological developments in PROMETHEE methods, stream Multiple Criteria Decision Aiding.
Monday, 14:30-16:00Room: Clarendon SR 2.08
Authors (first author is the speaker)
| 1. | Alexandre Flachs
|
| SMG - CoDE, Université libre de Bruxelles | |
| 2. | Yves De Smet
|
| SMG - CoDE, Université libre de Bruxelles |
Abstract
Promethee methods are a family of outranking methods from multiple criteria decision aid. These have been used in many practical applications and have been subject to multiple methodological extensions. Among Promethee methods, Promethee I and II allows building partial and complete rankings of the alternatives based on their evaluations regarding each criterion, using the positive, negative, and net flow scores. In the multiple criteria choice problem, a decision maker is given a set of alternatives evaluated regarding multiple conflicting criteria and is expected to select a subset of best alternatives corresponding to his/her preferences. One way to solve the choice problem is to build one or multiple rankings (depending on various instances of the preference parameters) of the alternatives and to choose the best subset by analyzing the obtained result. One can for example be interested in all alternatives that could be ranked first, with at least one instance of the method parameters. In this work we investigate under which conditions an alternative can be ranked first in a Promethee II ranking. Such questions were already raised in the context of Robust Ordinal Regression (ROR) with the concepts of possible and necessary ranks of an alternative. This contribution aims at being complementary to ROR by analyzing parameters requirements for an alternative to be ranked first, and draw conclusions on the expressivity of Promethee methods in terms of DM preferences.
Keywords
- Decision Support Systems
- Decision Analysis
- Decision Theory
Status: accepted
Back to the list of papers