Matches in SemOpenAlex for { <https://semopenalex.org/work/W2949008799> ?p ?o ?g. }
Showing items 1 to 95 of
95
with 100 items per page.
- W2949008799 abstract "Bayesian Dark Knowledge is a method for compressing the posterior predictive distribution of a neural network model into a more compact form. Specifically, the method attempts to compress a Monte Carlo approximation to the parameter posterior into a single network representing the posterior predictive distribution. Further, the authors show that this approach is successful in the classification setting using a student network whose architecture matches that of a single network in the teacher ensemble. In this work, we examine the robustness of Bayesian Dark Knowledge to higher levels of posterior uncertainty. We show that using a student network that matches the teacher architecture may fail to yield acceptable performance. We study an approach to close the resulting performance gap by increasing student model capacity." @default.
- W2949008799 created "2019-06-14" @default.
- W2949008799 creator A5023142437 @default.
- W2949008799 creator A5070743314 @default.
- W2949008799 date "2019-06-04" @default.
- W2949008799 modified "2023-09-27" @default.
- W2949008799 title "Assessing the Robustness of Bayesian Dark Knowledge to Posterior Uncertainty." @default.
- W2949008799 cites W1516111018 @default.
- W2949008799 cites W1567512734 @default.
- W2949008799 cites W1592410721 @default.
- W2949008799 cites W1719489212 @default.
- W2949008799 cites W2005708641 @default.
- W2949008799 cites W2047229728 @default.
- W2949008799 cites W2099471712 @default.
- W2949008799 cites W2108677974 @default.
- W2949008799 cites W2161758346 @default.
- W2949008799 cites W2167433878 @default.
- W2949008799 cites W2809702612 @default.
- W2949008799 cites W2995099985 @default.
- W2949008799 hasPublicationYear "2019" @default.
- W2949008799 type Work @default.
- W2949008799 sameAs 2949008799 @default.
- W2949008799 citedByCount "0" @default.
- W2949008799 crossrefType "posted-content" @default.
- W2949008799 hasAuthorship W2949008799A5023142437 @default.
- W2949008799 hasAuthorship W2949008799A5070743314 @default.
- W2949008799 hasConcept C104317684 @default.
- W2949008799 hasConcept C107673813 @default.
- W2949008799 hasConcept C119857082 @default.
- W2949008799 hasConcept C123657996 @default.
- W2949008799 hasConcept C124101348 @default.
- W2949008799 hasConcept C154945302 @default.
- W2949008799 hasConcept C160234255 @default.
- W2949008799 hasConcept C166957645 @default.
- W2949008799 hasConcept C185592680 @default.
- W2949008799 hasConcept C191413810 @default.
- W2949008799 hasConcept C193415008 @default.
- W2949008799 hasConcept C205649164 @default.
- W2949008799 hasConcept C33724603 @default.
- W2949008799 hasConcept C37903108 @default.
- W2949008799 hasConcept C38652104 @default.
- W2949008799 hasConcept C41008148 @default.
- W2949008799 hasConcept C50644808 @default.
- W2949008799 hasConcept C55493867 @default.
- W2949008799 hasConcept C57830394 @default.
- W2949008799 hasConcept C63479239 @default.
- W2949008799 hasConcept C83247935 @default.
- W2949008799 hasConceptScore W2949008799C104317684 @default.
- W2949008799 hasConceptScore W2949008799C107673813 @default.
- W2949008799 hasConceptScore W2949008799C119857082 @default.
- W2949008799 hasConceptScore W2949008799C123657996 @default.
- W2949008799 hasConceptScore W2949008799C124101348 @default.
- W2949008799 hasConceptScore W2949008799C154945302 @default.
- W2949008799 hasConceptScore W2949008799C160234255 @default.
- W2949008799 hasConceptScore W2949008799C166957645 @default.
- W2949008799 hasConceptScore W2949008799C185592680 @default.
- W2949008799 hasConceptScore W2949008799C191413810 @default.
- W2949008799 hasConceptScore W2949008799C193415008 @default.
- W2949008799 hasConceptScore W2949008799C205649164 @default.
- W2949008799 hasConceptScore W2949008799C33724603 @default.
- W2949008799 hasConceptScore W2949008799C37903108 @default.
- W2949008799 hasConceptScore W2949008799C38652104 @default.
- W2949008799 hasConceptScore W2949008799C41008148 @default.
- W2949008799 hasConceptScore W2949008799C50644808 @default.
- W2949008799 hasConceptScore W2949008799C55493867 @default.
- W2949008799 hasConceptScore W2949008799C57830394 @default.
- W2949008799 hasConceptScore W2949008799C63479239 @default.
- W2949008799 hasConceptScore W2949008799C83247935 @default.
- W2949008799 hasLocation W29490087991 @default.
- W2949008799 hasOpenAccess W2949008799 @default.
- W2949008799 hasPrimaryLocation W29490087991 @default.
- W2949008799 hasRelatedWork W2086486038 @default.
- W2949008799 hasRelatedWork W2137123230 @default.
- W2949008799 hasRelatedWork W2139650424 @default.
- W2949008799 hasRelatedWork W2165977312 @default.
- W2949008799 hasRelatedWork W2187318208 @default.
- W2949008799 hasRelatedWork W2293859833 @default.
- W2949008799 hasRelatedWork W2497473102 @default.
- W2949008799 hasRelatedWork W2558622672 @default.
- W2949008799 hasRelatedWork W2781494504 @default.
- W2949008799 hasRelatedWork W2910171846 @default.
- W2949008799 hasRelatedWork W2947972834 @default.
- W2949008799 hasRelatedWork W2963716487 @default.
- W2949008799 hasRelatedWork W2998352186 @default.
- W2949008799 hasRelatedWork W3033729636 @default.
- W2949008799 hasRelatedWork W3093945256 @default.
- W2949008799 hasRelatedWork W3110392801 @default.
- W2949008799 hasRelatedWork W3130438476 @default.
- W2949008799 hasRelatedWork W3132714944 @default.
- W2949008799 hasRelatedWork W3138077151 @default.
- W2949008799 hasRelatedWork W3192802934 @default.
- W2949008799 isParatext "false" @default.
- W2949008799 isRetracted "false" @default.
- W2949008799 magId "2949008799" @default.
- W2949008799 workType "article" @default.