Matches in SemOpenAlex for { <https://semopenalex.org/work/W3035689819> ?p ?o ?g. }
Showing items 1 to 98 of
98
with 100 items per page.
- W3035689819 abstract "Training deep neural networks (DNNs) in the presence of noisy labels is an important and challenging task. Probabilistic modeling, which consists of a classifier and a transition matrix, depicts the transformation from true labels to noisy labels and is a promising approach. However, recent probabilistic methods directly apply transition matrix to DNN, neglect DNN's susceptibility to overfitting, and achieve unsatisfactory performance, especially under the uniform noise. In this paper, inspired by label smoothing, we proposed a novel method, in which a smoothed transition matrix is used for updating DNN, to restrict the overfitting of DNN in probabilistic modeling. Our method is termed Matrix Smoothing. We also empirically demonstrate that our method not only improves the robustness of probabilistic modeling significantly, but also even obtains a better estimation of the transition matrix." @default.
- W3035689819 created "2020-06-19" @default.
- W3035689819 creator A5034104790 @default.
- W3035689819 creator A5052115680 @default.
- W3035689819 creator A5074930325 @default.
- W3035689819 date "2020-07-01" @default.
- W3035689819 modified "2023-09-28" @default.
- W3035689819 title "Matrix Smoothing: A Regularization For Dnn With Transition Matrix Under Noisy Labels" @default.
- W3035689819 cites W1821462560 @default.
- W3035689819 cites W1866072925 @default.
- W3035689819 cites W2108598243 @default.
- W3035689819 cites W2163605009 @default.
- W3035689819 cites W2183341477 @default.
- W3035689819 cites W2184045248 @default.
- W3035689819 cites W2194775991 @default.
- W3035689819 cites W2302255633 @default.
- W3035689819 cites W2403681572 @default.
- W3035689819 cites W2752971446 @default.
- W3035689819 cites W2891244534 @default.
- W3035689819 cites W2948210185 @default.
- W3035689819 cites W2963096987 @default.
- W3035689819 cites W2963403868 @default.
- W3035689819 cites W2963789034 @default.
- W3035689819 cites W2964155802 @default.
- W3035689819 cites W2964274690 @default.
- W3035689819 cites W2964292098 @default.
- W3035689819 cites W2979302305 @default.
- W3035689819 cites W2981873476 @default.
- W3035689819 cites W3137695714 @default.
- W3035689819 doi "https://doi.org/10.1109/icme46284.2020.9102853" @default.
- W3035689819 hasPublicationYear "2020" @default.
- W3035689819 type Work @default.
- W3035689819 sameAs 3035689819 @default.
- W3035689819 citedByCount "1" @default.
- W3035689819 countsByYear W30356898192020 @default.
- W3035689819 crossrefType "proceedings-article" @default.
- W3035689819 hasAuthorship W3035689819A5034104790 @default.
- W3035689819 hasAuthorship W3035689819A5052115680 @default.
- W3035689819 hasAuthorship W3035689819A5074930325 @default.
- W3035689819 hasBestOaLocation W30356898192 @default.
- W3035689819 hasConcept C104317684 @default.
- W3035689819 hasConcept C106487976 @default.
- W3035689819 hasConcept C11413529 @default.
- W3035689819 hasConcept C119857082 @default.
- W3035689819 hasConcept C153180895 @default.
- W3035689819 hasConcept C154945302 @default.
- W3035689819 hasConcept C159985019 @default.
- W3035689819 hasConcept C185592680 @default.
- W3035689819 hasConcept C192562407 @default.
- W3035689819 hasConcept C22019652 @default.
- W3035689819 hasConcept C2776135515 @default.
- W3035689819 hasConcept C31972630 @default.
- W3035689819 hasConcept C3770464 @default.
- W3035689819 hasConcept C41008148 @default.
- W3035689819 hasConcept C49555168 @default.
- W3035689819 hasConcept C49937458 @default.
- W3035689819 hasConcept C50644808 @default.
- W3035689819 hasConcept C55493867 @default.
- W3035689819 hasConcept C63479239 @default.
- W3035689819 hasConcept C98763669 @default.
- W3035689819 hasConceptScore W3035689819C104317684 @default.
- W3035689819 hasConceptScore W3035689819C106487976 @default.
- W3035689819 hasConceptScore W3035689819C11413529 @default.
- W3035689819 hasConceptScore W3035689819C119857082 @default.
- W3035689819 hasConceptScore W3035689819C153180895 @default.
- W3035689819 hasConceptScore W3035689819C154945302 @default.
- W3035689819 hasConceptScore W3035689819C159985019 @default.
- W3035689819 hasConceptScore W3035689819C185592680 @default.
- W3035689819 hasConceptScore W3035689819C192562407 @default.
- W3035689819 hasConceptScore W3035689819C22019652 @default.
- W3035689819 hasConceptScore W3035689819C2776135515 @default.
- W3035689819 hasConceptScore W3035689819C31972630 @default.
- W3035689819 hasConceptScore W3035689819C3770464 @default.
- W3035689819 hasConceptScore W3035689819C41008148 @default.
- W3035689819 hasConceptScore W3035689819C49555168 @default.
- W3035689819 hasConceptScore W3035689819C49937458 @default.
- W3035689819 hasConceptScore W3035689819C50644808 @default.
- W3035689819 hasConceptScore W3035689819C55493867 @default.
- W3035689819 hasConceptScore W3035689819C63479239 @default.
- W3035689819 hasConceptScore W3035689819C98763669 @default.
- W3035689819 hasLocation W30356898191 @default.
- W3035689819 hasLocation W30356898192 @default.
- W3035689819 hasOpenAccess W3035689819 @default.
- W3035689819 hasPrimaryLocation W30356898191 @default.
- W3035689819 hasRelatedWork W1996541855 @default.
- W3035689819 hasRelatedWork W2985459377 @default.
- W3035689819 hasRelatedWork W2989932438 @default.
- W3035689819 hasRelatedWork W3011996705 @default.
- W3035689819 hasRelatedWork W3099765033 @default.
- W3035689819 hasRelatedWork W3128220493 @default.
- W3035689819 hasRelatedWork W3175189414 @default.
- W3035689819 hasRelatedWork W3206592002 @default.
- W3035689819 hasRelatedWork W4210794429 @default.
- W3035689819 hasRelatedWork W4225691219 @default.
- W3035689819 isParatext "false" @default.
- W3035689819 isRetracted "false" @default.
- W3035689819 magId "3035689819" @default.
- W3035689819 workType "article" @default.