TY - GEN
T1 - On partially observable MDPs and BDI models
AU - Schut, Martijn
AU - Wooldridge, Michael
AU - Parsons, Simon
N1 - Publisher Copyright:
© Springer-Verlag Berlin Heidelberg 2002.
PY - 2002
Y1 - 2002
N2 - Decision theoretic planning in ai by means of solving Partially Observable Markov decision processes (pomdps) has been shown to be both powerful and versatile. However, such approaches are computationally hard and, from a design stance, are not necessarily intuitive for conceptualising many problems. We propose a novel method for solving pomdps, which provides a designer with a more intuitive means of specifying pomdp planning problems. In particular, we investigate the relationship between pomdp planning theory and belief-desire-intention (bdi) agent theory. The idea is to view a bdi agent as a specification of an pomdp problem. This view is to be supported by a correspondence between an pomdp problem and a bdi agent. In this paper, we outline such a correspondence between pomdp and bdi by explaining how to specify one in terms of the other. Additionally, we illustrate the significance of a correspondence by showing empirically that it yields satisfying results in complex domains.
AB - Decision theoretic planning in ai by means of solving Partially Observable Markov decision processes (pomdps) has been shown to be both powerful and versatile. However, such approaches are computationally hard and, from a design stance, are not necessarily intuitive for conceptualising many problems. We propose a novel method for solving pomdps, which provides a designer with a more intuitive means of specifying pomdp planning problems. In particular, we investigate the relationship between pomdp planning theory and belief-desire-intention (bdi) agent theory. The idea is to view a bdi agent as a specification of an pomdp problem. This view is to be supported by a correspondence between an pomdp problem and a bdi agent. In this paper, we outline such a correspondence between pomdp and bdi by explaining how to specify one in terms of the other. Additionally, we illustrate the significance of a correspondence by showing empirically that it yields satisfying results in complex domains.
UR - https://www.scopus.com/pages/publications/34247198718
U2 - 10.1007/3-540-45634-1_15
DO - 10.1007/3-540-45634-1_15
M3 - Conference contribution
SN - 3540439625
SN - 9783540439622
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 243
EP - 259
BT - Foundations and Applications of Multi-Agent Systems - UKMAS Workshops 1996-2000, Selected Papers
PB - Springer Verlag
T2 - UK Workshops on Multi-Agent Systems, UKMAS 1996-2000
Y2 - 1 January 2002
ER -