Matches in SemOpenAlex for { <https://semopenalex.org/work/W4288049759> ?p ?o ?g. }
Showing items 1 to 85 of
85
with 100 items per page.
- W4288049759 abstract "Mean rewards of actions are often correlated. The form of these correlations may be complex and unknown a priori, such as the preferences of a user for recommended products and their categories. To maximize statistical efficiency, it is important to leverage these correlations when learning. We formulate a bandit variant of this problem where the correlations of mean action rewards are represented by a hierarchical Bayesian model with latent variables. Since the hierarchy can have multiple layers, we call it deep. We propose a hierarchical Thompson sampling algorithm (HierTS) for this problem, and show how to implement it efficiently for Gaussian hierarchies. The efficient implementation is possible due to a novel exact hierarchical representation of the posterior, which itself is of independent interest. We use this exact posterior to analyze the Bayes regret of HierTS in Gaussian bandits. Our analysis reflects the structure of the problem, that the regret decreases with the prior width, and also shows that hierarchies reduce the regret by non-constant factors in the number of actions. We confirm these theoretical findings empirically, in both synthetic and real-world experiments." @default.
- W4288049759 created "2022-07-27" @default.
- W4288049759 creator A5005433975 @default.
- W4288049759 creator A5013843778 @default.
- W4288049759 creator A5018931712 @default.
- W4288049759 creator A5049020775 @default.
- W4288049759 creator A5065544942 @default.
- W4288049759 date "2022-02-03" @default.
- W4288049759 modified "2023-09-25" @default.
- W4288049759 title "Deep Hierarchy in Bandits" @default.
- W4288049759 doi "https://doi.org/10.48550/arxiv.2202.01454" @default.
- W4288049759 hasPublicationYear "2022" @default.
- W4288049759 type Work @default.
- W4288049759 citedByCount "0" @default.
- W4288049759 crossrefType "posted-content" @default.
- W4288049759 hasAuthorship W4288049759A5005433975 @default.
- W4288049759 hasAuthorship W4288049759A5013843778 @default.
- W4288049759 hasAuthorship W4288049759A5018931712 @default.
- W4288049759 hasAuthorship W4288049759A5049020775 @default.
- W4288049759 hasAuthorship W4288049759A5065544942 @default.
- W4288049759 hasBestOaLocation W42880497591 @default.
- W4288049759 hasConcept C107673813 @default.
- W4288049759 hasConcept C111472728 @default.
- W4288049759 hasConcept C119857082 @default.
- W4288049759 hasConcept C121332964 @default.
- W4288049759 hasConcept C124101348 @default.
- W4288049759 hasConcept C138885662 @default.
- W4288049759 hasConcept C144986985 @default.
- W4288049759 hasConcept C153083717 @default.
- W4288049759 hasConcept C154945302 @default.
- W4288049759 hasConcept C162324750 @default.
- W4288049759 hasConcept C163716315 @default.
- W4288049759 hasConcept C17744445 @default.
- W4288049759 hasConcept C199539241 @default.
- W4288049759 hasConcept C207201462 @default.
- W4288049759 hasConcept C2776359362 @default.
- W4288049759 hasConcept C31170391 @default.
- W4288049759 hasConcept C33923547 @default.
- W4288049759 hasConcept C34447519 @default.
- W4288049759 hasConcept C41008148 @default.
- W4288049759 hasConcept C50817715 @default.
- W4288049759 hasConcept C62520636 @default.
- W4288049759 hasConcept C73602740 @default.
- W4288049759 hasConcept C75553542 @default.
- W4288049759 hasConcept C94625758 @default.
- W4288049759 hasConceptScore W4288049759C107673813 @default.
- W4288049759 hasConceptScore W4288049759C111472728 @default.
- W4288049759 hasConceptScore W4288049759C119857082 @default.
- W4288049759 hasConceptScore W4288049759C121332964 @default.
- W4288049759 hasConceptScore W4288049759C124101348 @default.
- W4288049759 hasConceptScore W4288049759C138885662 @default.
- W4288049759 hasConceptScore W4288049759C144986985 @default.
- W4288049759 hasConceptScore W4288049759C153083717 @default.
- W4288049759 hasConceptScore W4288049759C154945302 @default.
- W4288049759 hasConceptScore W4288049759C162324750 @default.
- W4288049759 hasConceptScore W4288049759C163716315 @default.
- W4288049759 hasConceptScore W4288049759C17744445 @default.
- W4288049759 hasConceptScore W4288049759C199539241 @default.
- W4288049759 hasConceptScore W4288049759C207201462 @default.
- W4288049759 hasConceptScore W4288049759C2776359362 @default.
- W4288049759 hasConceptScore W4288049759C31170391 @default.
- W4288049759 hasConceptScore W4288049759C33923547 @default.
- W4288049759 hasConceptScore W4288049759C34447519 @default.
- W4288049759 hasConceptScore W4288049759C41008148 @default.
- W4288049759 hasConceptScore W4288049759C50817715 @default.
- W4288049759 hasConceptScore W4288049759C62520636 @default.
- W4288049759 hasConceptScore W4288049759C73602740 @default.
- W4288049759 hasConceptScore W4288049759C75553542 @default.
- W4288049759 hasConceptScore W4288049759C94625758 @default.
- W4288049759 hasLocation W42880497591 @default.
- W4288049759 hasOpenAccess W4288049759 @default.
- W4288049759 hasPrimaryLocation W42880497591 @default.
- W4288049759 hasRelatedWork W1600255059 @default.
- W4288049759 hasRelatedWork W2116720812 @default.
- W4288049759 hasRelatedWork W2890918712 @default.
- W4288049759 hasRelatedWork W2991918816 @default.
- W4288049759 hasRelatedWork W3128393330 @default.
- W4288049759 hasRelatedWork W3159366499 @default.
- W4288049759 hasRelatedWork W3166771275 @default.
- W4288049759 hasRelatedWork W3211418486 @default.
- W4288049759 hasRelatedWork W4300081859 @default.
- W4288049759 hasRelatedWork W4287329160 @default.
- W4288049759 isParatext "false" @default.
- W4288049759 isRetracted "false" @default.
- W4288049759 workType "article" @default.