Large-Scale Markov Decision Problems via the Linear Programming Dual
Abstract
We consider the problem of controlling a fully specified Markov decision process (MDP), also known as the planning problem, when the state space is very large and calculating the optimal policy is intractable. Instead, we pursue the more modest goal of optimizing over some small family of policies. Specifically, we show that the family of policies associated with a low-dimensional approximation of occupancy measures yields a tractable optimization. Moreover, we propose an efficient algorithm, scaling with the size of the subspace but not the state space, that is able to find a policy with low excess loss relative to the best policy in this class. To the best of our knowledge, such results did not exist in the literature previously. We bound excess loss in the average cost and discounted cost cases, which are treated separately. Preliminary experiments show the effectiveness of the proposed algorithms in a queueing application.
Cite
@article{arxiv.1901.01992,
title = {Large-Scale Markov Decision Problems via the Linear Programming Dual},
author = {Yasin Abbasi-Yadkori and Peter L. Bartlett and Xi Chen and Alan Malek},
journal= {arXiv preprint arXiv:1901.01992},
year = {2019}
}
Comments
53 pages. arXiv admin note: text overlap with arXiv:1402.6763