Markov Decision Processes and the Belief-Desire-Intention Model

Gerardo I. Simari Simon D. Parsons

ISBN 10: 1461414717 / ISBN 13: 9781461414711

Editore: Springer, 2011

Riguardo questo articolo

pp. 72.

Codice articolo 2648019735

Dati bibliografici

• Titolo: Markov Decision Processes and the Belief-Desire-Intention Model
• Editore: Springer
• Data di pubblicazione: 2011
• Lingua: Inglese
• ISBN 10: 1461414717
• ISBN 13: 9781461414711
• Rilegatura: Brossura
• Condizione: New

Descrizione articolo

Riassunto

In this work, we provide a treatment of the relationship between two models that have been widely used in the implementation of autonomous agents: the Belief DesireIntention (B.D.I.) model and Markov Decision Processes (M.D.P.s). We start with an informal description of the relationship, identifying the common features of the two approaches and the differences between them. Then we hone our understanding of these differences through an empirical analysis of the performance of both models on the TileWorld testbed. This allows us to show that even though the M.D.P. model displays consistently better behavior than the B.D.I./ model for small worlds, this is not the case when the world becomes large and the M.D.P. model cannot be solved exactly. Finally we present a theoretical analysis of the relationship between the two approaches, identifying mappings that allow us to extract a set of intentions from a policy (a solution to an M.D.P.), and to extract a policy from a set of intentions.

Contenuti

Introduction.- Preliminary Concepts.- An Empirical Comparison of Models.- A Theoretical Comparison of Models.- Related Work. Conclusions, Limitations, and Future Directions.