Matches in SemOpenAlex for { <https://semopenalex.org/work/W3128632247> ?p ?o ?g. }
- W3128632247 abstract "We propose a new probabilistic method for unsupervised recovery of corrupted data. Given a large ensemble of degraded samples, our method recovers accurate posteriors of clean values, allowing the exploration of the manifold of possible reconstructed data and hence characterising the underlying uncertainty. In this set-ting, direct application of classical variational methods often gives rise to collapsed densities that do not adequately explore the solution space. Instead, we derive our novel reduced entropy condition approximate inference method that results in rich posteriors. We test our model in a data recovery task under the common setting of missing values and noise, demonstrating superior performance to existing variational methods for imputation and de-noising with different real data sets. We further show higher classification accuracy after imputation, proving the advantage of propagating uncertainty to downstream tasks with our model." @default.
- W3128632247 created "2021-02-15" @default.
- W3128632247 creator A5017068912 @default.
- W3128632247 creator A5043183251 @default.
- W3128632247 creator A5070918132 @default.
- W3128632247 creator A5075176672 @default.
- W3128632247 date "2021-05-03" @default.
- W3128632247 modified "2023-09-28" @default.
- W3128632247 title "Tomographic Auto-Encoder: Unsupervised Bayesian Recovery of Corrupted Data" @default.
- W3128632247 cites W125693051 @default.
- W3128632247 cites W1857382374 @default.
- W3128632247 cites W1949839429 @default.
- W3128632247 cites W1966168183 @default.
- W3128632247 cites W1976086748 @default.
- W3128632247 cites W2007339694 @default.
- W3128632247 cites W2085411191 @default.
- W3128632247 cites W2108501770 @default.
- W3128632247 cites W2125389028 @default.
- W3128632247 cites W2129587342 @default.
- W3128632247 cites W2134843796 @default.
- W3128632247 cites W2155479981 @default.
- W3128632247 cites W2188365844 @default.
- W3128632247 cites W2437617937 @default.
- W3128632247 cites W2556707115 @default.
- W3128632247 cites W2557449848 @default.
- W3128632247 cites W2576683119 @default.
- W3128632247 cites W2744556021 @default.
- W3128632247 cites W2750384547 @default.
- W3128632247 cites W2786772298 @default.
- W3128632247 cites W2804184144 @default.
- W3128632247 cites W2885743790 @default.
- W3128632247 cites W2886436016 @default.
- W3128632247 cites W2886577208 @default.
- W3128632247 cites W2890686416 @default.
- W3128632247 cites W2894458036 @default.
- W3128632247 cites W2901093491 @default.
- W3128632247 cites W2902857081 @default.
- W3128632247 cites W2905258586 @default.
- W3128632247 cites W2920987997 @default.
- W3128632247 cites W2942471690 @default.
- W3128632247 cites W2950365845 @default.
- W3128632247 cites W2962908092 @default.
- W3128632247 cites W2962935371 @default.
- W3128632247 cites W2963073614 @default.
- W3128632247 cites W2963209089 @default.
- W3128632247 cites W2963258546 @default.
- W3128632247 cites W2963319176 @default.
- W3128632247 cites W2963501406 @default.
- W3128632247 cites W2963573392 @default.
- W3128632247 cites W2963925452 @default.
- W3128632247 cites W2964049407 @default.
- W3128632247 cites W2964204553 @default.
- W3128632247 cites W2964339842 @default.
- W3128632247 cites W3035623224 @default.
- W3128632247 hasPublicationYear "2021" @default.
- W3128632247 type Work @default.
- W3128632247 sameAs 3128632247 @default.
- W3128632247 citedByCount "0" @default.
- W3128632247 crossrefType "proceedings-article" @default.
- W3128632247 hasAuthorship W3128632247A5017068912 @default.
- W3128632247 hasAuthorship W3128632247A5043183251 @default.
- W3128632247 hasAuthorship W3128632247A5070918132 @default.
- W3128632247 hasAuthorship W3128632247A5075176672 @default.
- W3128632247 hasConcept C101738243 @default.
- W3128632247 hasConcept C106301342 @default.
- W3128632247 hasConcept C107673813 @default.
- W3128632247 hasConcept C108583219 @default.
- W3128632247 hasConcept C11413529 @default.
- W3128632247 hasConcept C119857082 @default.
- W3128632247 hasConcept C121332964 @default.
- W3128632247 hasConcept C124101348 @default.
- W3128632247 hasConcept C153180895 @default.
- W3128632247 hasConcept C154945302 @default.
- W3128632247 hasConcept C160920958 @default.
- W3128632247 hasConcept C2776214188 @default.
- W3128632247 hasConcept C41008148 @default.
- W3128632247 hasConcept C49937458 @default.
- W3128632247 hasConcept C58041806 @default.
- W3128632247 hasConcept C58489278 @default.
- W3128632247 hasConcept C62520636 @default.
- W3128632247 hasConcept C67186912 @default.
- W3128632247 hasConcept C77088390 @default.
- W3128632247 hasConcept C9357733 @default.
- W3128632247 hasConceptScore W3128632247C101738243 @default.
- W3128632247 hasConceptScore W3128632247C106301342 @default.
- W3128632247 hasConceptScore W3128632247C107673813 @default.
- W3128632247 hasConceptScore W3128632247C108583219 @default.
- W3128632247 hasConceptScore W3128632247C11413529 @default.
- W3128632247 hasConceptScore W3128632247C119857082 @default.
- W3128632247 hasConceptScore W3128632247C121332964 @default.
- W3128632247 hasConceptScore W3128632247C124101348 @default.
- W3128632247 hasConceptScore W3128632247C153180895 @default.
- W3128632247 hasConceptScore W3128632247C154945302 @default.
- W3128632247 hasConceptScore W3128632247C160920958 @default.
- W3128632247 hasConceptScore W3128632247C2776214188 @default.
- W3128632247 hasConceptScore W3128632247C41008148 @default.
- W3128632247 hasConceptScore W3128632247C49937458 @default.
- W3128632247 hasConceptScore W3128632247C58041806 @default.
- W3128632247 hasConceptScore W3128632247C58489278 @default.
- W3128632247 hasConceptScore W3128632247C62520636 @default.