Interval Iteration Algorithm for MDPs and IMDPs - Archive ouverte HAL Accéder directement au contenu
Article Dans Une Revue Theoretical Computer Science Année : 2018

Interval Iteration Algorithm for MDPs and IMDPs

Résumé

Markov Decision Processes (MDP) are a widely used model including both non-deterministic and probabilistic choices. Minimal and maximal probabilities to reach a target set of states, with respect to a policy resolving non-determinism, may be computed by several methods including value iteration. This algorithm, easy to implement and efficient in terms of space complexity, iteratively computes the probabilities of paths of increasing length. However, it raises three issues: (1) defining a stopping criterion ensuring a bound on the approximation, (2) analysing the rate of convergence, and (3) specifying an additional procedure to obtain the exact values once a sufficient number of iterations has been performed. The first two issues are still open and, for the third one, an upper bound on the number of iterations has been proposed. Based on a graph analysis and transformation of MDPs, we address these problems. First we introduce an interval iteration algorithm, for which the stopping criterion is straightforward. Then we exhibit its convergence rate. Finally we significantly improve the upper bound on the number of iterations required to get the exact values. We extend our approach to also deal with Interval Markov Decision Processes (IMDP) that can be seen as symbolic representations of MDPs.
Fichier principal
Vignette du fichier
tcs-version.pdf (488.24 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01809094 , version 1 (06-06-2018)

Identifiants

Citer

Serge Haddad, Benjamin Monmege. Interval Iteration Algorithm for MDPs and IMDPs. Theoretical Computer Science, 2018, 735, pp.111 - 131. ⟨10.1016/j.tcs.2016.12.003⟩. ⟨hal-01809094⟩
275 Consultations
879 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More