A novel individually rational objective in multi-agent multi-armed bandits: Algorithms and regret bounds
We study a two-player stochastic multi-armed bandit (MAB) problem with different expected rewards for each player, a generalisation of two-player general sum repeated games to stochastic rewards. Our aim is to find the egalitarian bargaining solution (EBS) for the repeated game, which can lead to mu...
Uloženo v:
| Vydáno v: | Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems Ročník 2020-May; s. 1395 |
|---|---|
| Hlavní autoři: | , , , |
| Médium: | Konferenční příspěvek |
| Jazyk: | angličtina |
| Vydáno: |
2020
|
| Témata: | |
| ISSN: | 1558-2914, 1548-8403 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Abstract | We study a two-player stochastic multi-armed bandit (MAB) problem with different expected rewards for each player, a generalisation of two-player general sum repeated games to stochastic rewards. Our aim is to find the egalitarian bargaining solution (EBS) for the repeated game, which can lead to much higher rewards than the maximin value of both players. Our main contribution is the derivation of an algorithm, UCRG, that achieves simultaneously for both players, a high-probability regret bound of order Õ ( T2/3) after any T rounds of play. We demonstrate that our upper bound is nearly optimal by proving a lower bound of Ω ( T2/3) for any algorithm. Experiments confirm our theoretical results and the superiority of UCRG compared to the well-known explore-then-commit heuristic. |
|---|---|
| AbstractList | We study a two-player stochastic multi-armed bandit (MAB) problem with different expected rewards for each player, a generalisation of two-player general sum repeated games to stochastic rewards. Our aim is to find the egalitarian bargaining solution (EBS) for the repeated game, which can lead to much higher rewards than the maximin value of both players. Our main contribution is the derivation of an algorithm, UCRG, that achieves simultaneously for both players, a high-probability regret bound of order Õ ( T2/3) after any T rounds of play. We demonstrate that our upper bound is nearly optimal by proving a lower bound of Ω ( T2/3) for any algorithm. Experiments confirm our theoretical results and the superiority of UCRG compared to the well-known explore-then-commit heuristic. |
| Author | Rzepecki, Jaroslaw Hofmann, K. Dimitrakakis, Christos Tossou, Aristide |
| Author_xml | – sequence: 1 givenname: Aristide surname: Tossou fullname: Tossou, Aristide organization: Data Science – sequence: 2 givenname: Christos surname: Dimitrakakis fullname: Dimitrakakis, Christos organization: University of Oslo – sequence: 3 givenname: Jaroslaw surname: Rzepecki fullname: Rzepecki, Jaroslaw organization: Microsoft Research – sequence: 4 givenname: K. surname: Hofmann fullname: Hofmann, K. organization: Microsoft Research |
| BackLink | https://research.chalmers.se/publication/521346$$DView record from Swedish Publication Index (Chalmers tekniska högskola) |
| BookMark | eNo1jMtOwzAURL0oEm1hzdY_kOJH7DjsqoqXVIkFsLauk-vGVR6VnRT17wkCZnNGM5pZkUU_9EjIHWcbNeteytIUmm9-WAqxIEuulMlEyfNrskrpyJjUJi-XpNnSfjhjS0Nfh3OoJ2jbC40whqGHlg7uiNUYzjj3tJvaMWRwwH7897HDmjqYt2N6oNv2MMQwNl2ic0QjHiKO1A1TX6cbcuWhTXj7xzX5fHr82L1k-7fn1912n4HgxmSqFL7wCr30Qkle1RJrDQwYV8wYx4TLHShUXuSVkkIxL5yqeSG157wwTq7J--9v-sLT5Owphg7ixQ4QbMSEEKvGVg20HcZkE1qhOFMloK18jTZnTlrQvLK-AHBaGy2Vkd-cyWsB |
| ContentType | Conference Proceeding |
| DBID | ADTPV BNKNJ F1S |
| DOI | 10.5555/3398761.3398922 |
| DatabaseName | SwePub SwePub Conference SWEPUB Chalmers tekniska högskola |
| DatabaseTitleList | |
| DeliveryMethod | fulltext_linktorsrc |
| Discipline | Computer Science |
| ExternalDocumentID | oai_research_chalmers_se_251059ae_cfde_40b3_a61c_f7aab6686358 |
| GroupedDBID | 123 29P 5VS 6IK 6IL AAJGR ADTPV ADZIZ ALMA_UNASSIGNED_HOLDINGS APO AVWKF BEFXN BFFAM BGNUA BKEBE BNKNJ BPEOZ CHZPO F1S I07 IPLJI OCL RIE RIL RNS |
| ID | FETCH-LOGICAL-a2188-592f7f5ef3f2531cd3ed6a0a015088b02b4ba5e5f24c53250f2b5d1736f1178b3 |
| ISSN | 1558-2914 1548-8403 |
| IngestDate | Wed Nov 05 04:22:36 EST 2025 |
| IsPeerReviewed | false |
| IsScholarly | false |
| Language | English |
| LinkModel | OpenURL |
| MergedId | FETCHMERGED-LOGICAL-a2188-592f7f5ef3f2531cd3ed6a0a015088b02b4ba5e5f24c53250f2b5d1736f1178b3 |
| ParticipantIDs | swepub_primary_oai_research_chalmers_se_251059ae_cfde_40b3_a61c_f7aab6686358 |
| PublicationCentury | 2000 |
| PublicationDate | 2020 |
| PublicationDateYYYYMMDD | 2020-01-01 |
| PublicationDate_xml | – year: 2020 text: 2020 |
| PublicationDecade | 2020 |
| PublicationTitle | Proceedings of the International Joint Conference on Autonomous Agents and Multiagent Systems |
| PublicationYear | 2020 |
| SSID | ssj0036849 |
| Score | 1.7420202 |
| Snippet | We study a two-player stochastic multi-armed bandit (MAB) problem with different expected rewards for each player, a generalisation of two-player general sum... |
| SourceID | swepub |
| SourceType | Open Access Repository |
| StartPage | 1395 |
| SubjectTerms | Egalitarian bargaining solution Individual rationality Multi-armed bandits Safety |
| Title | A novel individually rational objective in multi-agent multi-armed bandits: Algorithms and regret bounds |
| URI | https://research.chalmers.se/publication/521346 |
| Volume | 2020-May |
| hasFullText | 1 |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Lj9MwELbKwoETr0XAAvKBW5UliePE4VbxEIJltYci7S2yHbvt0iZVkla7_GJ-BmM7SUPpBQE9pK5bOWnm08x4Mt8MQq8i7qec59JLJOFexEnupSKCkaJgvoM4zbklCp8l5-fs8jK9GI1-dFyY7TIpCnZ9na7_q6hhDoRtqLN_IO5-UZiAMQgdjiB2OP4u-IP256KfrLsUgP3A36JoBmQ_88RgsmkMv8FkxE5mlvZmUzBMviE3n3-pbd5lOI-LcqtM1Y6O1bW8GVfdWUpx5bSpianYxEXPrdSOq5Xxfg2tpqnd04JZWS2a-cqdulKzSjVjYfo-9Y7_FGx6ubEKzWqnvMflO8PUqvg34xEPqibsUvm_q7WSrkn3Jw6-wXIQvi_1qm0W_fl0GAcJ_UEcpFXd1PBJHCW10-32h1_adCSnosHlpYdsB4UXyJaQFAxEcGreU0eZPlCQu63ENM_k3La5qbNaZaF1VrnKpM5VFvmCZDwOZKYTzkUcM_Dt2C10OyQwdEzDzmeAGbtRM_tHD7bclhbS_R9XkMpc3uu9i9srb2tdoul9dLzDD94h7gEaqeIhutd1C8Gt8XiE5hNs0YKHaMEdWnCPFvgeD9CCB2jBLVre4B1WMExhhxXssHKMvn54P3370Wv7fHgcHEzm0TTUiaZKEx2CSZA5UXnMfW6CcYwJPxSR4FRRHUYSNAj1dShoHiQk1kGQMEEeo6OiLNQThGmcUhIplgofbqMOmASXGPROyAkVMtFP0Zm7Y9naFXPJ_kqaz_7tcifo7g7Zz9FRU23UC3RHbptFXb20iPkJA4-5kg |
| linkProvider | IEEE |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=proceeding&rft.title=Proceedings+of+the+International+Joint+Conference+on+Autonomous+Agents+and+Multiagent+Systems&rft.atitle=A+novel+individually+rational+objective+in+multi-agent+multi-armed+bandits%3A+Algorithms+and+regret+bounds&rft.au=Tossou%2C+Aristide&rft.au=Dimitrakakis%2C+Christos&rft.au=Rzepecki%2C+Jaroslaw&rft.au=Hofmann%2C+K.&rft.date=2020-01-01&rft.issn=1558-2914&rft.volume=2020-May&rft.spage=1395&rft_id=info:doi/10.5555%2F3398761.3398922&rft.externalDocID=oai_research_chalmers_se_251059ae_cfde_40b3_a61c_f7aab6686358 |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1558-2914&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1558-2914&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1558-2914&client=summon |