FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3006322114> ?p ?o ?g. }

• W3006322114 abstract "Modern methods for Bayesian regression beyond the Gaussian response setting are computationally impractical or inaccurate in high dimensions. As discussed in recent literature, bypassing this trade-off is still an open problem even in basic binary regression models, and there is limited theory on the quality of variational approximations in high-dimensional settings. To address this gap, we study the approximation accuracy of routine-use mean-field variational Bayes in high-dimensional probit regression with Gaussian priors, obtaining new and practically relevant results on the pathological behavior of this strategy in uncertainty quantification, estimation and prediction, that also suggest caution against maximum a posteriori estimates when p>n. Motivated by these results, we develop a new partially-factorized variational approximation for the posterior distribution of the probit coefficients that leverages a representation with global and local variables but, unlike for classical mean-field assumptions, it avoids a fully factorized approximation, and instead assumes a factorization only for local variables. We prove that the resulting approximation belongs to a tractable class of unified skew-normal densities that incorporates skewness and, unlike for state-of-the-art mean-field solutions, converges to the exact posterior density as p goes to infinity. To solve the variational optimization problem, we derive a tractable CAVI algorithm that easily scales to p in tens of thousands, and provably requires a number of iterations converging to 1 as p goes to infinity. Such findings are also illustrated in extensive empirical studies where our new solution is shown to improve the accuracy of mean-field variational Bayes for any n and p, with the magnitude of these gains being remarkable in those high-dimensional p>n settings where state-of-the-art methods are computationally impractical." @default.
• W3006322114 created "2020-02-24" @default.
• W3006322114 creator A5025849454 @default.
• W3006322114 creator A5062491445 @default.
• W3006322114 creator A5068162496 @default.
• W3006322114 date "2019-11-15" @default.
• W3006322114 modified "2023-09-27" @default.
• W3006322114 title "Scalable and Accurate Variational Bayes for High-Dimensional Binary Regression Models" @default.
• W3006322114 cites W1487139063 @default.
• W3006322114 cites W1496451467 @default.
• W3006322114 cites W1905898145 @default.
• W3006322114 cites W1934021597 @default.
• W3006322114 cites W1965555277 @default.
• W3006322114 cites W1969902605 @default.
• W3006322114 cites W1979205968 @default.
• W3006322114 cites W1995780830 @default.
• W3006322114 cites W2006111398 @default.
• W3006322114 cites W2008703230 @default.
• W3006322114 cites W2041923730 @default.
• W3006322114 cites W2051378923 @default.
• W3006322114 cites W2051817458 @default.
• W3006322114 cites W2052517126 @default.
• W3006322114 cites W2061750459 @default.
• W3006322114 cites W2075337232 @default.
• W3006322114 cites W2078082181 @default.
• W3006322114 cites W2085893390 @default.
• W3006322114 cites W2094158759 @default.
• W3006322114 cites W2099878672 @default.
• W3006322114 cites W2108306139 @default.
• W3006322114 cites W2117853077 @default.
• W3006322114 cites W2131215891 @default.
• W3006322114 cites W2138429195 @default.
• W3006322114 cites W2139385756 @default.
• W3006322114 cites W2148319469 @default.
• W3006322114 cites W2149745547 @default.
• W3006322114 cites W2332757643 @default.
• W3006322114 cites W2796512382 @default.
• W3006322114 cites W2798428389 @default.
• W3006322114 cites W2887086773 @default.
• W3006322114 cites W2896691302 @default.
• W3006322114 cites W2899373970 @default.
• W3006322114 cites W2914815746 @default.
• W3006322114 cites W2963183953 @default.
• W3006322114 cites W2963254535 @default.
• W3006322114 cites W2963977107 @default.
• W3006322114 cites W2969557150 @default.
• W3006322114 cites W2982610605 @default.
• W3006322114 cites W2992769304 @default.
• W3006322114 cites W3029292097 @default.
• W3006322114 cites W3043651324 @default.
• W3006322114 cites W3084427384 @default.
• W3006322114 cites W3106380050 @default.
• W3006322114 cites W3125452207 @default.
• W3006322114 cites W621293842 @default.
• W3006322114 cites W639587122 @default.
• W3006322114 hasPublicationYear "2019" @default.
• W3006322114 type Work @default.
• W3006322114 sameAs 3006322114 @default.
• W3006322114 citedByCount "2" @default.
• W3006322114 countsByYear W30063221142020 @default.
• W3006322114 countsByYear W30063221142021 @default.
• W3006322114 crossrefType "posted-content" @default.
• W3006322114 hasAuthorship W3006322114A5025849454 @default.
• W3006322114 hasAuthorship W3006322114A5062491445 @default.
• W3006322114 hasAuthorship W3006322114A5068162496 @default.
• W3006322114 hasConcept C105795698 @default.
• W3006322114 hasConcept C107673813 @default.
• W3006322114 hasConcept C121332964 @default.
• W3006322114 hasConcept C126255220 @default.
• W3006322114 hasConcept C163716315 @default.
• W3006322114 hasConcept C177769412 @default.
• W3006322114 hasConcept C207201462 @default.
• W3006322114 hasConcept C28826006 @default.
• W3006322114 hasConcept C33923547 @default.
• W3006322114 hasConcept C41008148 @default.
• W3006322114 hasConcept C62520636 @default.
• W3006322114 hasConceptScore W3006322114C105795698 @default.
• W3006322114 hasConceptScore W3006322114C107673813 @default.
• W3006322114 hasConceptScore W3006322114C121332964 @default.
• W3006322114 hasConceptScore W3006322114C126255220 @default.
• W3006322114 hasConceptScore W3006322114C163716315 @default.
• W3006322114 hasConceptScore W3006322114C177769412 @default.
• W3006322114 hasConceptScore W3006322114C207201462 @default.
• W3006322114 hasConceptScore W3006322114C28826006 @default.
• W3006322114 hasConceptScore W3006322114C33923547 @default.
• W3006322114 hasConceptScore W3006322114C41008148 @default.
• W3006322114 hasConceptScore W3006322114C62520636 @default.
• W3006322114 hasLocation W30063221141 @default.
• W3006322114 hasOpenAccess W3006322114 @default.
• W3006322114 hasPrimaryLocation W30063221141 @default.
• W3006322114 hasRelatedWork W1567331820 @default.
• W3006322114 hasRelatedWork W2151454335 @default.
• W3006322114 hasRelatedWork W2170582592 @default.
• W3006322114 hasRelatedWork W2187471809 @default.
• W3006322114 hasRelatedWork W2235729179 @default.
• W3006322114 hasRelatedWork W2762763764 @default.
• W3006322114 hasRelatedWork W2765334337 @default.
• W3006322114 hasRelatedWork W2766152229 @default.
• W3006322114 hasRelatedWork W2883664406 @default.
• W3006322114 hasRelatedWork W2893924688 @default.

• next