Matches in SemOpenAlex for { <https://semopenalex.org/work/W2976996772> ?p ?o ?g. }
- W2976996772 abstract "This paper investigates a population-based training regime based on game-theoretic principles called Policy-Spaced Response Oracles (PSRO). PSRO is general in the sense that it (1) encompasses well-known algorithms such as fictitious play and double oracle as special cases, and (2) in principle applies to general-sum, many-player games. Despite this, prior studies of PSRO have been focused on two-player zero-sum games, a regime wherein Nash equilibria are tractably computable. In moving from two-player zero-sum games to more general settings, computation of Nash equilibria quickly becomes infeasible. Here, we extend the theoretical underpinnings of PSRO by considering an alternative solution concept, $alpha$-Rank, which is unique (thus faces no equilibrium selection issues, unlike Nash) and applies readily to general-sum, many-player settings. We establish convergence guarantees in several games classes, and identify links between Nash equilibria and $alpha$-Rank. We demonstrate the competitive performance of $alpha$-Rank-based PSRO against an exact Nash solver-based PSRO in 2-player Kuhn and Leduc Poker. We then go beyond the reach of prior PSRO applications by considering 3- to 5-player poker games, yielding instances where $alpha$-Rank achieves faster convergence than approximate Nash solvers, thus establishing it as a favorable general games solver. We also carry out an initial empirical validation in MuJoCo soccer, illustrating the feasibility of the proposed approach in another complex domain." @default.
- W2976996772 created "2019-10-03" @default.
- W2976996772 creator A5006533777 @default.
- W2976996772 creator A5006947993 @default.
- W2976996772 creator A5008547992 @default.
- W2976996772 creator A5031943811 @default.
- W2976996772 creator A5043984392 @default.
- W2976996772 creator A5045898861 @default.
- W2976996772 creator A5049659586 @default.
- W2976996772 creator A5051039278 @default.
- W2976996772 creator A5051619646 @default.
- W2976996772 creator A5052169592 @default.
- W2976996772 creator A5056053058 @default.
- W2976996772 creator A5056707583 @default.
- W2976996772 creator A5058922471 @default.
- W2976996772 creator A5062951341 @default.
- W2976996772 creator A5075375399 @default.
- W2976996772 date "2019-09-27" @default.
- W2976996772 modified "2023-09-23" @default.
- W2976996772 title "A Generalized Training Approach for Multiagent Learning." @default.
- W2976996772 cites W102212266 @default.
- W2976996772 cites W1192553058 @default.
- W2976996772 cites W1486118835 @default.
- W2976996772 cites W1486687115 @default.
- W2976996772 cites W1597864774 @default.
- W2976996772 cites W1759994832 @default.
- W2976996772 cites W1808543079 @default.
- W2976996772 cites W1988769349 @default.
- W2976996772 cites W2002373723 @default.
- W2976996772 cites W2023986907 @default.
- W2976996772 cites W2028798910 @default.
- W2976996772 cites W2067050450 @default.
- W2976996772 cites W2067768752 @default.
- W2976996772 cites W2082260760 @default.
- W2976996772 cites W2083535091 @default.
- W2976996772 cites W2090252037 @default.
- W2976996772 cites W2096145798 @default.
- W2976996772 cites W2104771847 @default.
- W2976996772 cites W2113351146 @default.
- W2976996772 cites W2126211987 @default.
- W2976996772 cites W2149254401 @default.
- W2976996772 cites W2152897361 @default.
- W2976996772 cites W2168356773 @default.
- W2976996772 cites W2291615277 @default.
- W2976996772 cites W2604873668 @default.
- W2976996772 cites W2604960219 @default.
- W2976996772 cites W2756196406 @default.
- W2976996772 cites W2768629321 @default.
- W2976996772 cites W2785324569 @default.
- W2976996772 cites W2902907165 @default.
- W2976996772 cites W2911616846 @default.
- W2976996772 cites W2925418831 @default.
- W2976996772 cites W2952465248 @default.
- W2976996772 cites W2962938168 @default.
- W2976996772 cites W2962966033 @default.
- W2976996772 cites W2963000099 @default.
- W2976996772 cites W2963048836 @default.
- W2976996772 cites W2963407617 @default.
- W2976996772 cites W2963485523 @default.
- W2976996772 cites W2963836708 @default.
- W2976996772 cites W2963937357 @default.
- W2976996772 cites W2964095117 @default.
- W2976996772 cites W2971037972 @default.
- W2976996772 cites W2981038142 @default.
- W2976996772 hasPublicationYear "2019" @default.
- W2976996772 type Work @default.
- W2976996772 sameAs 2976996772 @default.
- W2976996772 citedByCount "6" @default.
- W2976996772 countsByYear W29769967722019 @default.
- W2976996772 countsByYear W29769967722020 @default.
- W2976996772 countsByYear W29769967722021 @default.
- W2976996772 crossrefType "posted-content" @default.
- W2976996772 hasAuthorship W2976996772A5006533777 @default.
- W2976996772 hasAuthorship W2976996772A5006947993 @default.
- W2976996772 hasAuthorship W2976996772A5008547992 @default.
- W2976996772 hasAuthorship W2976996772A5031943811 @default.
- W2976996772 hasAuthorship W2976996772A5043984392 @default.
- W2976996772 hasAuthorship W2976996772A5045898861 @default.
- W2976996772 hasAuthorship W2976996772A5049659586 @default.
- W2976996772 hasAuthorship W2976996772A5051039278 @default.
- W2976996772 hasAuthorship W2976996772A5051619646 @default.
- W2976996772 hasAuthorship W2976996772A5052169592 @default.
- W2976996772 hasAuthorship W2976996772A5056053058 @default.
- W2976996772 hasAuthorship W2976996772A5056707583 @default.
- W2976996772 hasAuthorship W2976996772A5058922471 @default.
- W2976996772 hasAuthorship W2976996772A5062951341 @default.
- W2976996772 hasAuthorship W2976996772A5075375399 @default.
- W2976996772 hasConcept C114614502 @default.
- W2976996772 hasConcept C115903868 @default.
- W2976996772 hasConcept C126255220 @default.
- W2976996772 hasConcept C134306372 @default.
- W2976996772 hasConcept C144024400 @default.
- W2976996772 hasConcept C144237770 @default.
- W2976996772 hasConcept C149923435 @default.
- W2976996772 hasConcept C162324750 @default.
- W2976996772 hasConcept C164226766 @default.
- W2976996772 hasConcept C2777303404 @default.
- W2976996772 hasConcept C2778770139 @default.
- W2976996772 hasConcept C2908647359 @default.
- W2976996772 hasConcept C33923547 @default.