Efficient Policies for Stationary Possibilistic Markov Decision Processes

Nahla Ben Amor (1) , Zeineb El Khalfi (2) , Hélène Fargier (2, 3) , Régis Sabbadin (4)

1 LARODEC - Laboratoire de Recherche Opérationnelle de Décision et de Contrôle de Processus
2 IRIT-ADRIA - Argumentation, Décision, Raisonnement, Incertitude et Apprentissage
3 CNRS - Centre National de la Recherche Scientifique
4 MIAT INRA - Unité de Mathématiques et Informatique Appliquées de Toulouse

Possibilistic Markov Decision Processes offer a compact and tractable way to represent and solve problems of sequential decision under qualitative uncertainty. Even though appealing for its ability to handle qualitative problems, this model suffers from the drowning effect that is inherent to possibilistic decision theory. The present paper proposes to escape the drowning effect by extending to stationary possibilistic MDPs the lexicographic preference relations defined in [6] for non-sequential decision problems and provides a value iteration algorithm to compute policies that are optimal for these new criteria.

Mots clés: Possibility theory; Lexicographic comparisons; Possibilistic qualitative utilities; Markov Decision process

Domaines: Algorithme et structure de données [cs.DS]

Nahla Ben Amor, Zeineb El Khalfi, Hélène Fargier, Régis Sabbadin. Efficient Policies for Stationary Possibilistic Markov Decision Processes. ECSQARU 2017: European Conference on Symbolic and Quantitative Approaches to Reasoning with Uncertainty, Jul 2017, Lugano, Switzerland. pp.306-317, ⟨10.1007/978-3-319-61581-3_28⟩. ⟨hal-02863813⟩