Q-learning in regularized mean-field games
dc.contributor.author | Anahtarcı, Berkay | |
dc.contributor.author | Karıksız, Can Deha | |
dc.contributor.author | Saldı, N. | |
dc.date.accessioned | 2023-09-08T11:51:00Z | |
dc.date.available | 2023-09-08T11:51:00Z | |
dc.date.issued | 2023-03 | |
dc.identifier.issn | 2153-0785 | en_US |
dc.identifier.uri | http://hdl.handle.net/10679/8777 | |
dc.identifier.uri | https://link.springer.com/article/10.1007/s13235-022-00450-2 | |
dc.description.abstract | In this paper, we introduce a regularized mean-field game and study learning of this game under an infinite-horizon discounted reward function. Regularization is introduced by adding a strongly concave regularization function to the one-stage reward function in the classical mean-field game model. We establish a value iteration based learning algorithm to this regularized mean-field game using fitted Q-learning. The regularization term in general makes reinforcement learning algorithm more robust to the system components. Moreover, it enables us to establish error analysis of the learning algorithm without imposing restrictive convexity assumptions on the system components, which are needed in the absence of a regularization term. | en_US |
dc.description.sponsorship | BAGEP Award of the Science Academy | |
dc.language.iso | eng | en_US |
dc.publisher | Springer | en_US |
dc.relation.ispartof | Dynamic Games and Applications | |
dc.rights | restrictedAccess | |
dc.title | Q-learning in regularized mean-field games | en_US |
dc.type | Article | en_US |
dc.peerreviewed | yes | en_US |
dc.publicationstatus | Published | en_US |
dc.contributor.department | Özyeğin University | |
dc.contributor.authorID | (ORCID 0000-0001-6200-4398 & YÖK ID 331624) Anahtarcı, Berkay | |
dc.contributor.authorID | (ORCID 0000-0001-8890-2196 & YÖK ID ) Karıksız, Deha | |
dc.contributor.ozuauthor | Anahtarcı, Berkay | |
dc.contributor.ozuauthor | Karıksız, Can Deha | |
dc.identifier.volume | 13 | en_US |
dc.identifier.issue | 1 | en_US |
dc.identifier.startpage | 89 | en_US |
dc.identifier.endpage | 117 | en_US |
dc.identifier.wos | WOS:000800996400001 | |
dc.identifier.doi | 10.1007/s13235-022-00450-2 | en_US |
dc.subject.keywords | Discounted reward | en_US |
dc.subject.keywords | Mean-field games | en_US |
dc.subject.keywords | Q-learning | en_US |
dc.subject.keywords | Regularized Markov decision processes | en_US |
dc.identifier.scopus | SCOPUS:2-s2.0-85130543438 | |
dc.relation.publicationcategory | Article - International Refereed Journal - Institutional Academic Staff |
Files in this item
Files | Size | Format | View |
---|---|---|---|
There are no files associated with this item. |
This item appears in the following Collection(s)
Share this page