Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386494080> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W4386494080 abstract "We study how to learn $epsilon$-optimal strategies in zero-sum imperfect information games (IIG) with trajectory feedback. In this setting, players update their policies sequentially based on their observations over a fixed number of episodes, denoted by $T$. Existing procedures suffer from high variance due to the use of importance sampling over sequences of actions (Steinberger et al., 2020; McAleer et al., 2022). To reduce this variance, we consider a fixed sampling approach, where players still update their policies over time, but with observations obtained through a given fixed sampling policy. Our approach is based on an adaptive Online Mirror Descent (OMD) algorithm that applies OMD locally to each information set, using individually decreasing learning rates and a regularized loss. We show that this approach guarantees a convergence rate of $tilde{mathcal{O}}(T^{-1/2})$ with high probability and has a near-optimal dependence on the game parameters when applied with the best theoretical choices of learning rates and sampling policies. To achieve these results, we generalize the notion of OMD stabilization, allowing for time-varying regularization with convex increments." @default.
- W4386494080 created "2023-09-07" @default.
- W4386494080 creator A5006533777 @default.
- W4386494080 creator A5016746833 @default.
- W4386494080 creator A5037297959 @default.
- W4386494080 creator A5067550667 @default.
- W4386494080 creator A5070075141 @default.
- W4386494080 creator A5070500506 @default.
- W4386494080 date "2023-09-01" @default.
- W4386494080 modified "2023-10-16" @default.
- W4386494080 title "Local and adaptive mirror descents in extensive-form games" @default.
- W4386494080 doi "https://doi.org/10.48550/arxiv.2309.00656" @default.
- W4386494080 hasPublicationYear "2023" @default.
- W4386494080 type Work @default.
- W4386494080 citedByCount "0" @default.
- W4386494080 crossrefType "posted-content" @default.
- W4386494080 hasAuthorship W4386494080A5006533777 @default.
- W4386494080 hasAuthorship W4386494080A5016746833 @default.
- W4386494080 hasAuthorship W4386494080A5037297959 @default.
- W4386494080 hasAuthorship W4386494080A5067550667 @default.
- W4386494080 hasAuthorship W4386494080A5070075141 @default.
- W4386494080 hasAuthorship W4386494080A5070500506 @default.
- W4386494080 hasBestOaLocation W43864940801 @default.
- W4386494080 hasConcept C106131492 @default.
- W4386494080 hasConcept C112680207 @default.
- W4386494080 hasConcept C118615104 @default.
- W4386494080 hasConcept C121955636 @default.
- W4386494080 hasConcept C123676819 @default.
- W4386494080 hasConcept C126255220 @default.
- W4386494080 hasConcept C127162648 @default.
- W4386494080 hasConcept C138885662 @default.
- W4386494080 hasConcept C140779682 @default.
- W4386494080 hasConcept C144237770 @default.
- W4386494080 hasConcept C154945302 @default.
- W4386494080 hasConcept C162324750 @default.
- W4386494080 hasConcept C177264268 @default.
- W4386494080 hasConcept C196083921 @default.
- W4386494080 hasConcept C199360897 @default.
- W4386494080 hasConcept C2524010 @default.
- W4386494080 hasConcept C2776135515 @default.
- W4386494080 hasConcept C2780310539 @default.
- W4386494080 hasConcept C31258907 @default.
- W4386494080 hasConcept C31972630 @default.
- W4386494080 hasConcept C33923547 @default.
- W4386494080 hasConcept C36686422 @default.
- W4386494080 hasConcept C41008148 @default.
- W4386494080 hasConcept C41895202 @default.
- W4386494080 hasConcept C57869625 @default.
- W4386494080 hasConceptScore W4386494080C106131492 @default.
- W4386494080 hasConceptScore W4386494080C112680207 @default.
- W4386494080 hasConceptScore W4386494080C118615104 @default.
- W4386494080 hasConceptScore W4386494080C121955636 @default.
- W4386494080 hasConceptScore W4386494080C123676819 @default.
- W4386494080 hasConceptScore W4386494080C126255220 @default.
- W4386494080 hasConceptScore W4386494080C127162648 @default.
- W4386494080 hasConceptScore W4386494080C138885662 @default.
- W4386494080 hasConceptScore W4386494080C140779682 @default.
- W4386494080 hasConceptScore W4386494080C144237770 @default.
- W4386494080 hasConceptScore W4386494080C154945302 @default.
- W4386494080 hasConceptScore W4386494080C162324750 @default.
- W4386494080 hasConceptScore W4386494080C177264268 @default.
- W4386494080 hasConceptScore W4386494080C196083921 @default.
- W4386494080 hasConceptScore W4386494080C199360897 @default.
- W4386494080 hasConceptScore W4386494080C2524010 @default.
- W4386494080 hasConceptScore W4386494080C2776135515 @default.
- W4386494080 hasConceptScore W4386494080C2780310539 @default.
- W4386494080 hasConceptScore W4386494080C31258907 @default.
- W4386494080 hasConceptScore W4386494080C31972630 @default.
- W4386494080 hasConceptScore W4386494080C33923547 @default.
- W4386494080 hasConceptScore W4386494080C36686422 @default.
- W4386494080 hasConceptScore W4386494080C41008148 @default.
- W4386494080 hasConceptScore W4386494080C41895202 @default.
- W4386494080 hasConceptScore W4386494080C57869625 @default.
- W4386494080 hasLocation W43864940801 @default.
- W4386494080 hasOpenAccess W4386494080 @default.
- W4386494080 hasPrimaryLocation W43864940801 @default.
- W4386494080 hasRelatedWork W1974907279 @default.
- W4386494080 hasRelatedWork W1986454804 @default.
- W4386494080 hasRelatedWork W2040689219 @default.
- W4386494080 hasRelatedWork W2218420294 @default.
- W4386494080 hasRelatedWork W2620037948 @default.
- W4386494080 hasRelatedWork W3121672722 @default.
- W4386494080 hasRelatedWork W3122255429 @default.
- W4386494080 hasRelatedWork W3125866650 @default.
- W4386494080 hasRelatedWork W4298257953 @default.
- W4386494080 hasRelatedWork W937639825 @default.
- W4386494080 isParatext "false" @default.
- W4386494080 isRetracted "false" @default.
- W4386494080 workType "article" @default.