Matches in SemOpenAlex for { <https://semopenalex.org/work/W3169452765> ?p ?o ?g. }
- W3169452765 abstract "This paper investigates the problem of computing the equilibrium of competitive games, which is often modeled as a constrained saddle-point optimization problem with probability simplex constraints. Despite recent efforts in understanding the last-iterate convergence of extragradient methods in the unconstrained setting, the theoretical underpinnings of these methods in the constrained settings, especially those using multiplicative updates, remain highly inadequate, even when the objective function is bilinear. Motivated by the algorithmic role of entropy regularization in single-agent reinforcement learning and game theory, we develop provably efficient extragradient methods to find the quantal response equilibrium (QRE) -- which are solutions to zero-sum two-player matrix games with entropy regularization -- at a linear rate. The proposed algorithms can be implemented in a decentralized manner, where each player executes symmetric and multiplicative updates iteratively using its own payoff without observing the opponent's actions directly. In addition, by controlling the knob of entropy regularization, the proposed algorithms can locate an approximate Nash equilibrium of the unregularized matrix game at a sublinear rate without assuming the Nash equilibrium to be unique. Our methods also lead to efficient policy extragradient algorithms for solving (entropy-regularized) zero-sum Markov games at similar rates. All of our convergence rates are nearly dimension-free, which are independent of the size of the state and action spaces up to logarithm factors, highlighting the positive role of entropy regularization for accelerating convergence." @default.
- W3169452765 created "2021-06-22" @default.
- W3169452765 creator A5005015806 @default.
- W3169452765 creator A5053809095 @default.
- W3169452765 creator A5091389636 @default.
- W3169452765 date "2021-05-31" @default.
- W3169452765 modified "2023-09-23" @default.
- W3169452765 title "Fast Policy Extragradient Methods for Competitive Games with Entropy Regularization" @default.
- W3169452765 cites W1496590343 @default.
- W3169452765 cites W1519983590 @default.
- W3169452765 cites W1542941925 @default.
- W3169452765 cites W1570963478 @default.
- W3169452765 cites W1710476689 @default.
- W3169452765 cites W2048027382 @default.
- W3169452765 cites W2075226379 @default.
- W3169452765 cites W2075567596 @default.
- W3169452765 cites W2089559088 @default.
- W3169452765 cites W2093246563 @default.
- W3169452765 cites W2096913736 @default.
- W3169452765 cites W2106887613 @default.
- W3169452765 cites W2150865801 @default.
- W3169452765 cites W2156737235 @default.
- W3169452765 cites W2254533881 @default.
- W3169452765 cites W2264897026 @default.
- W3169452765 cites W2575731723 @default.
- W3169452765 cites W2619268125 @default.
- W3169452765 cites W2863175877 @default.
- W3169452765 cites W2894677249 @default.
- W3169452765 cites W2914920107 @default.
- W3169452765 cites W2951445759 @default.
- W3169452765 cites W2960066928 @default.
- W3169452765 cites W2963297691 @default.
- W3169452765 cites W2963508732 @default.
- W3169452765 cites W2963647223 @default.
- W3169452765 cites W2964067523 @default.
- W3169452765 cites W2964070557 @default.
- W3169452765 cites W2964334726 @default.
- W3169452765 cites W2971302301 @default.
- W3169452765 cites W3010842970 @default.
- W3169452765 cites W3034426742 @default.
- W3169452765 cites W3035454135 @default.
- W3169452765 cites W3037593317 @default.
- W3169452765 cites W3041970508 @default.
- W3169452765 cites W3046553904 @default.
- W3169452765 cites W3095423866 @default.
- W3169452765 cites W3098384494 @default.
- W3169452765 cites W3100292177 @default.
- W3169452765 cites W3106398159 @default.
- W3169452765 cites W3127686539 @default.
- W3169452765 cites W3127950324 @default.
- W3169452765 cites W3131948096 @default.
- W3169452765 cites W3131996280 @default.
- W3169452765 cites W3159062549 @default.
- W3169452765 cites W3164106810 @default.
- W3169452765 cites W361876 @default.
- W3169452765 cites W3034039613 @default.
- W3169452765 doi "https://doi.org/10.48550/arxiv.2105.15186" @default.
- W3169452765 hasPublicationYear "2021" @default.
- W3169452765 type Work @default.
- W3169452765 sameAs 3169452765 @default.
- W3169452765 citedByCount "4" @default.
- W3169452765 countsByYear W31694527652021 @default.
- W3169452765 crossrefType "posted-content" @default.
- W3169452765 hasAuthorship W3169452765A5005015806 @default.
- W3169452765 hasAuthorship W3169452765A5053809095 @default.
- W3169452765 hasAuthorship W3169452765A5091389636 @default.
- W3169452765 hasBestOaLocation W31694527651 @default.
- W3169452765 hasConcept C106301342 @default.
- W3169452765 hasConcept C117160843 @default.
- W3169452765 hasConcept C118615104 @default.
- W3169452765 hasConcept C121332964 @default.
- W3169452765 hasConcept C126255220 @default.
- W3169452765 hasConcept C127162648 @default.
- W3169452765 hasConcept C134306372 @default.
- W3169452765 hasConcept C136356330 @default.
- W3169452765 hasConcept C144237770 @default.
- W3169452765 hasConcept C154945302 @default.
- W3169452765 hasConcept C22171661 @default.
- W3169452765 hasConcept C2776135515 @default.
- W3169452765 hasConcept C28826006 @default.
- W3169452765 hasConcept C31258907 @default.
- W3169452765 hasConcept C33923547 @default.
- W3169452765 hasConcept C39927690 @default.
- W3169452765 hasConcept C41008148 @default.
- W3169452765 hasConcept C42747912 @default.
- W3169452765 hasConcept C46814582 @default.
- W3169452765 hasConcept C57869625 @default.
- W3169452765 hasConcept C62520636 @default.
- W3169452765 hasConceptScore W3169452765C106301342 @default.
- W3169452765 hasConceptScore W3169452765C117160843 @default.
- W3169452765 hasConceptScore W3169452765C118615104 @default.
- W3169452765 hasConceptScore W3169452765C121332964 @default.
- W3169452765 hasConceptScore W3169452765C126255220 @default.
- W3169452765 hasConceptScore W3169452765C127162648 @default.
- W3169452765 hasConceptScore W3169452765C134306372 @default.
- W3169452765 hasConceptScore W3169452765C136356330 @default.
- W3169452765 hasConceptScore W3169452765C144237770 @default.
- W3169452765 hasConceptScore W3169452765C154945302 @default.
- W3169452765 hasConceptScore W3169452765C22171661 @default.
- W3169452765 hasConceptScore W3169452765C2776135515 @default.