English

Randomized Linear Programming Solves the Discounted Markov Decision Problem In Nearly-Linear (Sometimes Sublinear) Running Time

Optimization and Control 2019-06-04 v3 Data Structures and Algorithms

Abstract

We propose a novel randomized linear programming algorithm for approximating the optimal policy of the discounted Markov decision problem. By leveraging the value-policy duality and binary-tree data structures, the algorithm adaptively samples state-action-state transitions and makes exponentiated primal-dual updates. We show that it finds an ϵ\epsilon-optimal policy using nearly-linear run time in the worst case. When the Markov decision process is ergodic and specified in some special data formats, the algorithm finds an ϵ\epsilon-optimal policy using run time linear in the total number of state-action pairs, which is sublinear in the input size. These results provide a new venue and complexity benchmarks for solving stochastic dynamic programs.

Keywords

Cite

@article{arxiv.1704.01869,
  title  = {Randomized Linear Programming Solves the Discounted Markov Decision Problem In Nearly-Linear (Sometimes Sublinear) Running Time},
  author = {Mengdi Wang},
  journal= {arXiv preprint arXiv:1704.01869},
  year   = {2019}
}
R2 v1 2026-06-22T19:09:47.853Z