Matches in SemOpenAlex for { <https://semopenalex.org/work/W2978833282> ?p ?o ?g. }
Showing items 1 to 100 of
100
with 100 items per page.
- W2978833282 abstract "In this work we consider the problem of learning a classifier from noisy labels when a few labeled examples are given. The structure of and noisy data is modeled by a graph per class and Graph Convolutional Networks (GCN) are used to predict class relevance of noisy examples. For each class, the GCN is treated as a binary classifier, which learns to discriminate from noisy examples using a weighted binary cross-entropy loss function. The GCN-inferred clean probability is then exploited as a relevance measure. Each noisy example is weighted by its relevance when learning a classifier for the end task. We evaluate our method on an extended version of a few-shot learning problem, where the few examples of novel classes are supplemented with additional noisy data. Experimental results show that our GCN-based cleaning process significantly improves the classification accuracy over not cleaning the noisy data, as well as standard few-shot classification where only few examples are used." @default.
- W2978833282 created "2019-10-10" @default.
- W2978833282 creator A5019409717 @default.
- W2978833282 creator A5030078277 @default.
- W2978833282 creator A5045217258 @default.
- W2978833282 creator A5046083819 @default.
- W2978833282 creator A5088574130 @default.
- W2978833282 date "2019-11-19" @default.
- W2978833282 modified "2023-09-27" @default.
- W2978833282 title "Graph Convolutional Networks for Learning with Few Clean and many Noisy Labels" @default.
- W2978833282 cites W1514928307 @default.
- W2978833282 cites W1665214252 @default.
- W2978833282 cites W1866072925 @default.
- W2978833282 cites W2100031962 @default.
- W2978833282 cites W2113290770 @default.
- W2978833282 cites W2117539524 @default.
- W2978833282 cites W2131791003 @default.
- W2978833282 cites W2145680191 @default.
- W2978833282 cites W2151575489 @default.
- W2978833282 cites W2152790380 @default.
- W2978833282 cites W2154455818 @default.
- W2978833282 cites W2159291644 @default.
- W2978833282 cites W219040644 @default.
- W2978833282 cites W2194775991 @default.
- W2978833282 cites W2250384498 @default.
- W2978833282 cites W2407712691 @default.
- W2978833282 cites W2558748708 @default.
- W2978833282 cites W2559091987 @default.
- W2978833282 cites W2577784528 @default.
- W2978833282 cites W2592335154 @default.
- W2978833282 cites W2601450892 @default.
- W2978833282 cites W2604763608 @default.
- W2978833282 cites W2732026016 @default.
- W2978833282 cites W2753160622 @default.
- W2978833282 cites W2770468159 @default.
- W2978833282 cites W2796346823 @default.
- W2978833282 cites W2798836702 @default.
- W2978833282 cites W2798991696 @default.
- W2978833282 cites W2895671740 @default.
- W2978833282 cites W2962722184 @default.
- W2978833282 cites W2962762068 @default.
- W2978833282 cites W2962762541 @default.
- W2978833282 cites W2962895018 @default.
- W2978833282 cites W2963078860 @default.
- W2978833282 cites W2963263347 @default.
- W2978833282 cites W2963341924 @default.
- W2978833282 cites W2963371670 @default.
- W2978833282 cites W2963741406 @default.
- W2978833282 cites W2963845150 @default.
- W2978833282 cites W2964015378 @default.
- W2978833282 cites W2964026991 @default.
- W2978833282 cites W2964078140 @default.
- W2978833282 cites W2964155802 @default.
- W2978833282 cites W2964292098 @default.
- W2978833282 cites W2964311892 @default.
- W2978833282 cites W2964321699 @default.
- W2978833282 cites W637153065 @default.
- W2978833282 hasPublicationYear "2019" @default.
- W2978833282 type Work @default.
- W2978833282 sameAs 2978833282 @default.
- W2978833282 citedByCount "2" @default.
- W2978833282 countsByYear W29788332822020 @default.
- W2978833282 countsByYear W29788332822021 @default.
- W2978833282 crossrefType "posted-content" @default.
- W2978833282 hasAuthorship W2978833282A5019409717 @default.
- W2978833282 hasAuthorship W2978833282A5030078277 @default.
- W2978833282 hasAuthorship W2978833282A5045217258 @default.
- W2978833282 hasAuthorship W2978833282A5046083819 @default.
- W2978833282 hasAuthorship W2978833282A5088574130 @default.
- W2978833282 hasBestOaLocation W29788332821 @default.
- W2978833282 hasConcept C132525143 @default.
- W2978833282 hasConcept C154945302 @default.
- W2978833282 hasConcept C41008148 @default.
- W2978833282 hasConcept C80444323 @default.
- W2978833282 hasConcept C81363708 @default.
- W2978833282 hasConceptScore W2978833282C132525143 @default.
- W2978833282 hasConceptScore W2978833282C154945302 @default.
- W2978833282 hasConceptScore W2978833282C41008148 @default.
- W2978833282 hasConceptScore W2978833282C80444323 @default.
- W2978833282 hasConceptScore W2978833282C81363708 @default.
- W2978833282 hasLocation W29788332821 @default.
- W2978833282 hasLocation W29788332822 @default.
- W2978833282 hasLocation W29788332823 @default.
- W2978833282 hasLocation W29788332824 @default.
- W2978833282 hasOpenAccess W2978833282 @default.
- W2978833282 hasPrimaryLocation W29788332821 @default.
- W2978833282 hasRelatedWork W2521062615 @default.
- W2978833282 hasRelatedWork W2735477435 @default.
- W2978833282 hasRelatedWork W2748454020 @default.
- W2978833282 hasRelatedWork W2749468216 @default.
- W2978833282 hasRelatedWork W2901465038 @default.
- W2978833282 hasRelatedWork W2998526951 @default.
- W2978833282 hasRelatedWork W3090822330 @default.
- W2978833282 hasRelatedWork W3119610945 @default.
- W2978833282 hasRelatedWork W3181746755 @default.
- W2978833282 hasRelatedWork W4239686595 @default.
- W2978833282 isParatext "false" @default.
- W2978833282 isRetracted "false" @default.
- W2978833282 magId "2978833282" @default.
- W2978833282 workType "article" @default.