Matches in SemOpenAlex for { <https://semopenalex.org/work/W4288804560> ?p ?o ?g. }
Showing items 1 to 71 of
71
with 100 items per page.
- W4288804560 abstract "An abundance of recent impossibility results establish that regret minimization in Markov games with adversarial opponents is both statistically and computationally intractable. Nevertheless, none of these results preclude the possibility of regret minimization under the assumption that all parties adopt the same learning procedure. In this work, we present the first (to our knowledge) algorithm for learning in general-sum Markov games that provides sublinear regret guarantees when executed by all agents. The bounds we obtain are for swap regret, and thus, along the way, imply convergence to a correlated equilibrium. Our algorithm is decentralized, computationally efficient, and does not require any communication between agents. Our key observation is that online learning via policy optimization in Markov games essentially reduces to a form of weighted regret minimization, with unknown weights determined by the path length of the agents' policy sequence. Consequently, controlling the path length leads to weighted regret objectives for which sufficiently adaptive algorithms provide sublinear regret guarantees." @default.
- W4288804560 created "2022-07-30" @default.
- W4288804560 creator A5000561336 @default.
- W4288804560 creator A5014148612 @default.
- W4288804560 creator A5014637159 @default.
- W4288804560 creator A5037346243 @default.
- W4288804560 creator A5053775980 @default.
- W4288804560 date "2022-07-28" @default.
- W4288804560 modified "2023-10-14" @default.
- W4288804560 title "Regret Minimization and Convergence to Equilibria in General-sum Markov Games" @default.
- W4288804560 doi "https://doi.org/10.48550/arxiv.2207.14211" @default.
- W4288804560 hasPublicationYear "2022" @default.
- W4288804560 type Work @default.
- W4288804560 citedByCount "0" @default.
- W4288804560 crossrefType "posted-content" @default.
- W4288804560 hasAuthorship W4288804560A5000561336 @default.
- W4288804560 hasAuthorship W4288804560A5014148612 @default.
- W4288804560 hasAuthorship W4288804560A5014637159 @default.
- W4288804560 hasAuthorship W4288804560A5037346243 @default.
- W4288804560 hasAuthorship W4288804560A5053775980 @default.
- W4288804560 hasBestOaLocation W42888045601 @default.
- W4288804560 hasConcept C105795698 @default.
- W4288804560 hasConcept C106189395 @default.
- W4288804560 hasConcept C117160843 @default.
- W4288804560 hasConcept C118615104 @default.
- W4288804560 hasConcept C119857082 @default.
- W4288804560 hasConcept C126255220 @default.
- W4288804560 hasConcept C144237770 @default.
- W4288804560 hasConcept C159886148 @default.
- W4288804560 hasConcept C162324750 @default.
- W4288804560 hasConcept C199360897 @default.
- W4288804560 hasConcept C2777303404 @default.
- W4288804560 hasConcept C2777735758 @default.
- W4288804560 hasConcept C33923547 @default.
- W4288804560 hasConcept C41008148 @default.
- W4288804560 hasConcept C50522688 @default.
- W4288804560 hasConcept C50817715 @default.
- W4288804560 hasConcept C98763669 @default.
- W4288804560 hasConceptScore W4288804560C105795698 @default.
- W4288804560 hasConceptScore W4288804560C106189395 @default.
- W4288804560 hasConceptScore W4288804560C117160843 @default.
- W4288804560 hasConceptScore W4288804560C118615104 @default.
- W4288804560 hasConceptScore W4288804560C119857082 @default.
- W4288804560 hasConceptScore W4288804560C126255220 @default.
- W4288804560 hasConceptScore W4288804560C144237770 @default.
- W4288804560 hasConceptScore W4288804560C159886148 @default.
- W4288804560 hasConceptScore W4288804560C162324750 @default.
- W4288804560 hasConceptScore W4288804560C199360897 @default.
- W4288804560 hasConceptScore W4288804560C2777303404 @default.
- W4288804560 hasConceptScore W4288804560C2777735758 @default.
- W4288804560 hasConceptScore W4288804560C33923547 @default.
- W4288804560 hasConceptScore W4288804560C41008148 @default.
- W4288804560 hasConceptScore W4288804560C50522688 @default.
- W4288804560 hasConceptScore W4288804560C50817715 @default.
- W4288804560 hasConceptScore W4288804560C98763669 @default.
- W4288804560 hasLocation W42888045601 @default.
- W4288804560 hasOpenAccess W4288804560 @default.
- W4288804560 hasPrimaryLocation W42888045601 @default.
- W4288804560 hasRelatedWork W10227384 @default.
- W4288804560 hasRelatedWork W11809405 @default.
- W4288804560 hasRelatedWork W1826788 @default.
- W4288804560 hasRelatedWork W2191283 @default.
- W4288804560 hasRelatedWork W4776762 @default.
- W4288804560 hasRelatedWork W491107 @default.
- W4288804560 hasRelatedWork W5133103 @default.
- W4288804560 hasRelatedWork W5718419 @default.
- W4288804560 hasRelatedWork W8540740 @default.
- W4288804560 hasRelatedWork W9932698 @default.
- W4288804560 isParatext "false" @default.
- W4288804560 isRetracted "false" @default.
- W4288804560 workType "article" @default.