Publications

• Sorted by Date • Classified by Publication Type • Classified by Research Category •

Efficient Exploitation of Factored Domains in Bayesian Reinforcement Learning for POMDPs

Sammie Katt, Frans A. Oliehoek, and Christopher Amato. Efficient Exploitation of Factored Domains in Bayesian Reinforcement Learning for POMDPs. In Proceedings of the AAMAS Workshop on Adaptive Learning Agents (ALA), July 2018.

Download

pdf [826.1kB]  

Abstract

While the POMDP has proven to be a powerful framework to modeland solve partially observable stochastic problems, it assumes ac-curate and complete knowledge of the environment. When suchinformation is not available, as is the case in many real world appli-cations, one must learn such a model. The BA-POMDP considersthe model as part of the hidden state and explicitly considers theuncertainty over it, and as a result transforms the learning probleminto a planning problem. This model, however, grows exponentiallywith the underlying POMDP size, and becomes intractable for non-trivial problems. In this article we propose a factored framework,the FBA-POMDP that represents the model as a Bayes-Net, dras-tically decreasing the number of parameters required to describethe dynamics of the environment. We demonstrate that the our ap-proach allows solvers to tackle problems much larger than possiblein the BA-POMDP.

BibTeX Entry

@inproceedings{Katt18ALA,
    title =     {Efficient Exploitation of Factored Domains in Bayesian
                Reinforcement Learning for {POMDPs}},
    author =    {Sammie Katt and Frans A. Oliehoek and Christopher Amato},
    booktitle = ALA18,
    year =     2018,
    month =     jul,
    url =       {http://ala2018.it.nuigalway.ie/papers/ALA\_2018\_paper\_49.pdf},
    keywords =   {refereed, workshop},
    abstract={
While the POMDP has proven to be a powerful framework to model
and solve partially observable stochastic problems, it assumes ac-
curate and complete knowledge of the environment. When such
information is not available, as is the case in many real world appli-
cations, one must learn such a model. The BA-POMDP considers
the model as part of the hidden state and explicitly considers the
uncertainty over it, and as a result transforms the learning problem
into a planning problem. This model, however, grows exponentially
with the underlying POMDP size, and becomes intractable for non-
trivial problems. In this article we propose a factored framework,
the FBA-POMDP that represents the model as a Bayes-Net, dras-
tically decreasing the number of parameters required to describe
the dynamics of the environment. We demonstrate that the our ap-
proach allows solvers to tackle problems much larger than possible
in the BA-POMDP.        
    }
}

Generated by bib2html.pl (written by Patrick Riley) on Tue Nov 05, 2024 16:13:37 UTC