On Value Iteration Convergence in Connected MDPs
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911916788547584 |
|---|---|
| author | Mustafin, Arsenii Olshevsky, Alex Paschalidis, Ioannis Ch. |
| author_facet | Mustafin, Arsenii Olshevsky, Alex Paschalidis, Ioannis Ch. |
| contents | 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 γ for both discounted and average-reward criteria. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_09592 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | On Value Iteration Convergence in Connected MDPs Mustafin, Arsenii Olshevsky, Alex Paschalidis, Ioannis Ch. Machine Learning 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 γ for both discounted and average-reward criteria. |
| title | On Value Iteration Convergence in Connected MDPs |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2406.09592 |