An adaptive sampling algorithm for solving Markov decision processes

An adaptive sampling algorithm for solving Markov decision processes

0.00 Avg rating0 Votes
Article ID: iaor20073850
Country: United States
Volume: 53
Issue: 1
Start Page Number: 126
End Page Number: 139
Publication Date: Jan 2005
Journal: Operations Research
Authors: , , ,
Keywords: programming: dynamic, control processes
Abstract:

Based on recent results for multiarmed bandit problems, we propose an adaptive sampling algorithm that approximates the optimal value of a finite-horizon Markov decision process (MDP) with finite state and action spaces. The algorithm adaptively chooses which action to sample as the sampling process proceeds and generates an asymptotically unbiased estimator, whose bias is bounded by a quantity that converges to zero at rate (ln N)/N, where N is the total number of samples that are used per state sampled in each stage. The worst-case running-time complexity of the algorithm is O((|A|N)H), independent of the size of the state space, where |A| is the size of the action space and H is the horizon length. The algorithm can be used to create an approximate receding horizon control to solve infinite-horizon MDPs. To illustrate the algorithm, computational results are reported on simple examples from inventory control.

Reviews

Required fields are marked *. Your email address will not be published.