English

Markov decision processes: on the convergence of the Monte-Carlo first visit algorithm

Probability 2025-09-23 v2 Optimization and Control

Abstract

We consider the Monte-Carlo first visit algorithm, of which the goal is to find the optimal control in a Markov decision process with finite state space and finite number of possible actions. We show its convergence when the discount factor is smaller than 1/21/2.

Keywords

Cite

@article{arxiv.2501.08800,
  title  = {Markov decision processes: on the convergence of the Monte-Carlo first visit algorithm},
  author = {Sylvain Delattre and Nicolas Fournier},
  journal= {arXiv preprint arXiv:2501.08800},
  year   = {2025}
}
R2 v1 2026-06-28T21:07:10.129Z