AIFeb 6, 2013
Region-Based Approximations for Planning in Stochastic Domains
arXiv:1302.1573v18 citations
Originality Incremental advance
AI Analysis
This addresses the problem of planning in stochastic domains for AI and robotics researchers, but it appears incremental as it builds on existing POMDP methods.
The paper tackles the difficulty of solving partially observable Markov decision processes (POMDPs) by identifying a subclass called region observable POMDPs, which are easier to solve and can approximate general POMDPs to arbitrary accuracy.
This paper is concerned with planning in stochastic domains by means of partially observable Markov decision processes (POMDPs). POMDPs are difficult to solve. This paper identifies a subclass of POMDPs called region observable POMDPs, which are easier to solve and can be used to approximate general POMDPs to arbitrary accuracy.