TY - JOUR
T1 - Optimal control of infinite horizon partially observable decision processes modelled as generators of probabilistic regular languages
AU - Chattopadhyay, Ishanu
AU - Ray, Asok
N1 - Funding Information:
This work has been supported in part by the US Army Research Laboratory (ARL) and the US Army Research Office (ARO) under Grant No. W911NF-07-1-0376, by the US Office of Naval Research (ONR) under Grant No. N00014-09-1-0688, and by NASA under Cooperative Agreement No. NNX07AK49A. Any opinions, findings and conclusions or recommendations expressed in this publication are those of the authors and do not necessarily reflect the views of the sponsoring agencies.
PY - 2010/3
Y1 - 2010/3
N2 - Decision processes with incomplete state feedback have been traditionally modelled as partially observable Markov decision processes. In this article, we present an alternative formulation based on probabilistic regular languages. The proposed approach generalises the recently reported work on language measure theoretic optimal control for perfectly observable situations and shows that such a framework is far more computationally tractable to the classical alternative. In particular, we show that the infinite horizon decision problem under partial observation, modelled in the proposed framework, is λ-approximable and, in general, is not harder to solve compared to the fully observable case. The approach is illustrated via two simple examples.
AB - Decision processes with incomplete state feedback have been traditionally modelled as partially observable Markov decision processes. In this article, we present an alternative formulation based on probabilistic regular languages. The proposed approach generalises the recently reported work on language measure theoretic optimal control for perfectly observable situations and shows that such a framework is far more computationally tractable to the classical alternative. In particular, we show that the infinite horizon decision problem under partial observation, modelled in the proposed framework, is λ-approximable and, in general, is not harder to solve compared to the fully observable case. The approach is illustrated via two simple examples.
UR - http://www.scopus.com/inward/record.url?scp=77951115539&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77951115539&partnerID=8YFLogxK
U2 - 10.1080/00207170903193458
DO - 10.1080/00207170903193458
M3 - Article
AN - SCOPUS:77951115539
SN - 0020-7179
VL - 83
SP - 457
EP - 483
JO - International Journal of Control
JF - International Journal of Control
IS - 3
ER -