Cautious OWA and Evidential Reasoning for decision under uncertainty

J.M. Tacnet (1) , J. Dezert (2)

1 UR ETGR (ETNA) - Érosion torrentielle, neige et avalanches
2 ONERA - The French Aerospace Lab [Palaiseau]

To make a decision under certainty, multicriteria decision methods aims to choose, rank or sort alternatives on the basis of quantitative or qualitative criteria and preferences expressed by the decision-makers. However, decision is often done under uncertainty: choosing alternatives can have different consequences depending on the external context (or state of the world). In this paper, a new methodology called Cautious Ordered Weighted Averaging with Evidential Reasoning (COWA-ER) is proposed for decision making under uncertainty to take into account imperfect evaluations of the alternatives and unknown beliefs about groups of the possible states of the world (scenarii). COWA-ER mixes cautiously the principle of Yager's Ordered Weighted Averaging (OWA) approach with the efficient fusion of belief functions proposed in Dezert-Smarandache Theory (DSmT).

Mots clés: INCERTITUDE; AIDE A LA DECISION; ANALYSE MULTICRITERE; FUSION D'INFORMATION; ORDERED WEIGHTED AVERAGING; DEZERT SMARANDACHE THEORY

Domaines: Sciences de l'environnement

J.M. Tacnet, J. Dezert. Cautious OWA and Evidential Reasoning for decision under uncertainty. 14th International Conference on Information Fusion, Jul 2011, Chicago, United States. 8 p. ⟨hal-00647164⟩