Value-Function Approximations for Partially Observable Markov Decision Processes
M. Hauskrecht. Journal of Artificial Intelligence Research, (2000)
Аннотация
Partially observable Markov decision processes (POMDPs) provide an elegant mathematical
framework for modeling complex decision and planning problems in stochastic
domains in which states of the system are observable only indirectly, via a set of imperfect
or noisy observations. The modeling advantage of POMDPs, however, comes at a price |
exact methods for solving them are computationally very expensive and thus applicable
in practice only to very simple problems. We focus on ecient...
%0 Journal Article
%1 citeulike:519644
%A Hauskrecht, Milos
%D 2000
%J Journal of Artificial Intelligence Research
%K grid-interpolation pomdp
%P 33--94
%T Value-Function Approximations for Partially Observable Markov Decision Processes
%U http://citeseer.ist.psu.edu/hauskrecht00valuefunction.html
%V 13
%X Partially observable Markov decision processes (POMDPs) provide an elegant mathematical
framework for modeling complex decision and planning problems in stochastic
domains in which states of the system are observable only indirectly, via a set of imperfect
or noisy observations. The modeling advantage of POMDPs, however, comes at a price |
exact methods for solving them are computationally very expensive and thus applicable
in practice only to very simple problems. We focus on ecient...
@article{citeulike:519644,
abstract = {Partially observable Markov decision processes (POMDPs) provide an elegant mathematical
framework for modeling complex decision and planning problems in stochastic
domains in which states of the system are observable only indirectly, via a set of imperfect
or noisy observations. The modeling advantage of POMDPs, however, comes at a price |
exact methods for solving them are computationally very expensive and thus applicable
in practice only to very simple problems. We focus on ecient...},
added-at = {2006-04-12T20:53:54.000+0200},
author = {Hauskrecht, Milos},
biburl = {https://www.bibsonomy.org/bibtex/2c7554d64602a51d041188691d933ce21/darius},
citeulike-article-id = {519644},
description = {CiteULike},
interhash = {aa674f14e36c1743a5261ea967a55517},
intrahash = {c7554d64602a51d041188691d933ce21},
journal = {Journal of Artificial Intelligence Research},
keywords = {grid-interpolation pomdp},
pages = {33--94},
priority = {2},
timestamp = {2006-04-12T20:53:54.000+0200},
title = {Value-Function Approximations for Partially Observable Markov Decision Processes},
url = {http://citeseer.ist.psu.edu/hauskrecht00valuefunction.html},
volume = 13,
year = 2000
}