Region-Based Approximations for Planning in Stochastic Domains

Abstract

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.

Appears in Proceedings of the Thirteenth Conference on Uncertainty in Artificial Intelligence (UAI1997)

Cited by in corpus (2)

• Value-Function Approximations for Partially Observable Markov Decision Processes
• My Brain is Full: When More Memory Helps