Matches in SemOpenAlex for { <https://semopenalex.org/work/W2166566250> ?p ?o ?g. }
- W2166566250 endingPage "3265" @default.
- W2166566250 startingPage "3250" @default.
- W2166566250 abstract "Many applications require optimizing an unknown, noisy function that is expensive to evaluate. We formalize this task as a multi-armed bandit problem, where the payoff function is either sampled from a Gaussian process (GP) or has low RKHS norm. We resolve the important open problem of deriving regret bounds for this setting, which imply novel convergence rates for GP optimization. We analyze GP-UCB, an intuitive upper-confidence based algorithm, and bound its cumulative regret in terms of maximal information gain, establishing a novel connection between GP optimization and experimental design. Moreover, by bounding the latter in terms of operator spectra, we obtain explicit sublinear regret bounds for many commonly used covariance functions. In some important cases, our bounds have surprisingly weak dependence on the dimensionality. In our experiments on real sensor data, GP-UCB compares favorably with other heuristical GP optimization approaches." @default.
- W2166566250 created "2016-06-24" @default.
- W2166566250 creator A5003040843 @default.
- W2166566250 creator A5005681375 @default.
- W2166566250 creator A5009319251 @default.
- W2166566250 creator A5018792915 @default.
- W2166566250 date "2012-05-01" @default.
- W2166566250 modified "2023-10-18" @default.
- W2166566250 title "Information-Theoretic Regret Bounds for Gaussian Process Optimization in the Bandit Setting" @default.
- W2166566250 cites W1510052597 @default.
- W2166566250 cites W1680189815 @default.
- W2166566250 cites W1998498767 @default.
- W2166566250 cites W2014018052 @default.
- W2166566250 cites W2063180182 @default.
- W2166566250 cites W2078011655 @default.
- W2166566250 cites W2083459869 @default.
- W2166566250 cites W2098642746 @default.
- W2166566250 cites W2099111195 @default.
- W2166566250 cites W2115519224 @default.
- W2166566250 cites W2133255100 @default.
- W2166566250 cites W2139095575 @default.
- W2166566250 cites W2168405694 @default.
- W2166566250 cites W3098799846 @default.
- W2166566250 cites W4236876623 @default.
- W2166566250 cites W4298876635 @default.
- W2166566250 cites W55912154 @default.
- W2166566250 doi "https://doi.org/10.1109/tit.2011.2182033" @default.
- W2166566250 hasPublicationYear "2012" @default.
- W2166566250 type Work @default.
- W2166566250 sameAs 2166566250 @default.
- W2166566250 citedByCount "607" @default.
- W2166566250 countsByYear W21665662502012 @default.
- W2166566250 countsByYear W21665662502013 @default.
- W2166566250 countsByYear W21665662502014 @default.
- W2166566250 countsByYear W21665662502015 @default.
- W2166566250 countsByYear W21665662502016 @default.
- W2166566250 countsByYear W21665662502017 @default.
- W2166566250 countsByYear W21665662502018 @default.
- W2166566250 countsByYear W21665662502019 @default.
- W2166566250 countsByYear W21665662502020 @default.
- W2166566250 countsByYear W21665662502021 @default.
- W2166566250 countsByYear W21665662502022 @default.
- W2166566250 countsByYear W21665662502023 @default.
- W2166566250 crossrefType "journal-article" @default.
- W2166566250 hasAuthorship W2166566250A5003040843 @default.
- W2166566250 hasAuthorship W2166566250A5005681375 @default.
- W2166566250 hasAuthorship W2166566250A5009319251 @default.
- W2166566250 hasAuthorship W2166566250A5018792915 @default.
- W2166566250 hasBestOaLocation W21665662502 @default.
- W2166566250 hasConcept C105795698 @default.
- W2166566250 hasConcept C111030470 @default.
- W2166566250 hasConcept C117160843 @default.
- W2166566250 hasConcept C118615104 @default.
- W2166566250 hasConcept C119857082 @default.
- W2166566250 hasConcept C121332964 @default.
- W2166566250 hasConcept C126255220 @default.
- W2166566250 hasConcept C134306372 @default.
- W2166566250 hasConcept C137836250 @default.
- W2166566250 hasConcept C14036430 @default.
- W2166566250 hasConcept C154945302 @default.
- W2166566250 hasConcept C163716315 @default.
- W2166566250 hasConcept C178650346 @default.
- W2166566250 hasConcept C33923547 @default.
- W2166566250 hasConcept C41008148 @default.
- W2166566250 hasConcept C50817715 @default.
- W2166566250 hasConcept C61326573 @default.
- W2166566250 hasConcept C62520636 @default.
- W2166566250 hasConcept C63584917 @default.
- W2166566250 hasConcept C77553402 @default.
- W2166566250 hasConcept C78458016 @default.
- W2166566250 hasConcept C86803240 @default.
- W2166566250 hasConceptScore W2166566250C105795698 @default.
- W2166566250 hasConceptScore W2166566250C111030470 @default.
- W2166566250 hasConceptScore W2166566250C117160843 @default.
- W2166566250 hasConceptScore W2166566250C118615104 @default.
- W2166566250 hasConceptScore W2166566250C119857082 @default.
- W2166566250 hasConceptScore W2166566250C121332964 @default.
- W2166566250 hasConceptScore W2166566250C126255220 @default.
- W2166566250 hasConceptScore W2166566250C134306372 @default.
- W2166566250 hasConceptScore W2166566250C137836250 @default.
- W2166566250 hasConceptScore W2166566250C14036430 @default.
- W2166566250 hasConceptScore W2166566250C154945302 @default.
- W2166566250 hasConceptScore W2166566250C163716315 @default.
- W2166566250 hasConceptScore W2166566250C178650346 @default.
- W2166566250 hasConceptScore W2166566250C33923547 @default.
- W2166566250 hasConceptScore W2166566250C41008148 @default.
- W2166566250 hasConceptScore W2166566250C50817715 @default.
- W2166566250 hasConceptScore W2166566250C61326573 @default.
- W2166566250 hasConceptScore W2166566250C62520636 @default.
- W2166566250 hasConceptScore W2166566250C63584917 @default.
- W2166566250 hasConceptScore W2166566250C77553402 @default.
- W2166566250 hasConceptScore W2166566250C78458016 @default.
- W2166566250 hasConceptScore W2166566250C86803240 @default.
- W2166566250 hasIssue "5" @default.
- W2166566250 hasLocation W21665662501 @default.
- W2166566250 hasLocation W21665662502 @default.
- W2166566250 hasLocation W21665662503 @default.
- W2166566250 hasOpenAccess W2166566250 @default.