Büchi objectives in countable MDPs
<p>We study countably infinite Markov decision processes with Büchi objectives, which ask to visit a given subset F of states infinitely often. A question left open by T.P. Hill in 1979 [10] is whether there always exist ε-optimal Markov strategies, i.e., strategies that base decisions only on...
| Main Authors: | , , , |
|---|---|
| Format: | Conference item |
| Published: |
Schloss Dagstuhl
2019
|
| _version_ | 1797096177262395392 |
|---|---|
| author | Kiefer, S Mayr, R Shirmohammadi, M Totzke, P |
| author_facet | Kiefer, S Mayr, R Shirmohammadi, M Totzke, P |
| author_sort | Kiefer, S |
| collection | OXFORD |
| description | <p>We study countably infinite Markov decision processes with Büchi objectives, which ask to visit a given subset F of states infinitely often. A question left open by T.P. Hill in 1979 [10] is whether there always exist ε-optimal Markov strategies, i.e., strategies that base decisions only on the current state and the number of steps taken so far. We provide a negative answer to this question by constructing a non-trivial counterexample. On the other hand, we show that Markov strategies with only 1 bit of extra memory are sufficient.</p> |
| first_indexed | 2024-03-07T04:38:21Z |
| format | Conference item |
| id | oxford-uuid:d0c1b9b9-35b0-4905-a3bf-06fa75430b5c |
| institution | University of Oxford |
| last_indexed | 2024-03-07T04:38:21Z |
| publishDate | 2019 |
| publisher | Schloss Dagstuhl |
| record_format | dspace |
| spelling | oxford-uuid:d0c1b9b9-35b0-4905-a3bf-06fa75430b5c2022-03-27T07:52:14ZBüchi objectives in countable MDPsConference itemhttp://purl.org/coar/resource_type/c_5794uuid:d0c1b9b9-35b0-4905-a3bf-06fa75430b5cSymplectic Elements at OxfordSchloss Dagstuhl2019Kiefer, SMayr, RShirmohammadi, MTotzke, P<p>We study countably infinite Markov decision processes with Büchi objectives, which ask to visit a given subset F of states infinitely often. A question left open by T.P. Hill in 1979 [10] is whether there always exist ε-optimal Markov strategies, i.e., strategies that base decisions only on the current state and the number of steps taken so far. We provide a negative answer to this question by constructing a non-trivial counterexample. On the other hand, we show that Markov strategies with only 1 bit of extra memory are sufficient.</p> |
| spellingShingle | Kiefer, S Mayr, R Shirmohammadi, M Totzke, P Büchi objectives in countable MDPs |
| title | Büchi objectives in countable MDPs |
| title_full | Büchi objectives in countable MDPs |
| title_fullStr | Büchi objectives in countable MDPs |
| title_full_unstemmed | Büchi objectives in countable MDPs |
| title_short | Büchi objectives in countable MDPs |
| title_sort | buchi objectives in countable mdps |
| work_keys_str_mv | AT kiefers buchiobjectivesincountablemdps AT mayrr buchiobjectivesincountablemdps AT shirmohammadim buchiobjectivesincountablemdps AT totzkep buchiobjectivesincountablemdps |