Work

Model-based Bayesian reinforcement learning in partially observable domains

Year: 2008

Type: conference-paper

Abstract: Bayesian reinforcement learning in partially observable domains is notoriously difficult, in part due to the unknown form of the beliefs and the optimal value function.We show that beliefs represented by mixtures of products of Dirichlet distributions are closed under belief updates for factored domains.Belief monitoring algorithms that use this mixture representation are proposed.We also show that the optimal value function is a linear combination of products of Dirichlets for factored domains.Finally, we extend BEETLE, which is a point-based value iteration algorithm for Bayesian RL in fully observable domains, to partially observable domains.

Source: Open Repository and Bibliography (University of Luxembourg)

Authors: Pascal Poupart, Nikos Vlassis

Institutions: University of Luxembourg, Luxembourg Centre for Systems Biomedicine

Language: English

Topic: Reinforcement Learning in Robotics

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

OpenAlex ID: W2158640361

FWCI: 7.115

Cites: 19

Cited by: 80

Related to: 10