English

On Value Iteration Convergence in Connected MDPs

Machine Learning 2024-06-17 v1

Abstract

This paper establishes that an MDP with a unique optimal policy and ergodic associated transition matrix ensures the convergence of various versions of the Value Iteration algorithm at a geometric rate that exceeds the discount factor {\gamma} for both discounted and average-reward criteria.

Keywords

Cite

@article{arxiv.2406.09592,
  title  = {On Value Iteration Convergence in Connected MDPs},
  author = {Arsenii Mustafin and Alex Olshevsky and Ioannis Ch. Paschalidis},
  journal= {arXiv preprint arXiv:2406.09592},
  year   = {2024}
}

Comments

8 pages, 1 figure

R2 v1 2026-06-28T17:05:19.898Z