Matches in SemOpenAlex for { <https://semopenalex.org/work/W3176879439> ?p ?o ?g. }
Showing items 1 to 70 of
70
with 100 items per page.
- W3176879439 endingPage "15868" @default.
- W3176879439 startingPage "15867" @default.
- W3176879439 abstract "Multi-task learning (MTL) is frequently used in settings where a target task has to be learnt based on limited training data, but knowledge can be leveraged from related auxiliary tasks. While MTL can improve task performance overall relative to single-task learning (STL), these improvements can hide negative transfer (NT), where STL may deliver better performance for many individual tasks. Asymmetric multi-task feature learning (AMTFL) is an approach that tries to address this by allowing tasks with higher loss values to have smaller influence on feature representations for learning other tasks. Task loss values do not necessarily indicate reliability of models for a specific task. We present examples of NT in two orthogonal datasets (image recognition and pharmacogenomics) and tackle this challenge by using aleatoric homoscedastic uncertainty to capture the relative confidence between tasks, and set weights for task loss. Our results show that this approach reduces NT providing a new approach to enable robust MTL." @default.
- W3176879439 created "2021-07-05" @default.
- W3176879439 creator A5001546652 @default.
- W3176879439 creator A5058488564 @default.
- W3176879439 creator A5062774997 @default.
- W3176879439 date "2021-05-18" @default.
- W3176879439 modified "2023-09-25" @default.
- W3176879439 title "Task Uncertainty Loss Reduce Negative Transfer in Asymmetric Multi-task Feature Learning (Student Abstract)" @default.
- W3176879439 doi "https://doi.org/10.1609/aaai.v35i18.17930" @default.
- W3176879439 hasPublicationYear "2021" @default.
- W3176879439 type Work @default.
- W3176879439 sameAs 3176879439 @default.
- W3176879439 citedByCount "1" @default.
- W3176879439 countsByYear W31768794392023 @default.
- W3176879439 crossrefType "journal-article" @default.
- W3176879439 hasAuthorship W3176879439A5001546652 @default.
- W3176879439 hasAuthorship W3176879439A5058488564 @default.
- W3176879439 hasAuthorship W3176879439A5062774997 @default.
- W3176879439 hasBestOaLocation W31768794391 @default.
- W3176879439 hasConcept C101104100 @default.
- W3176879439 hasConcept C104409967 @default.
- W3176879439 hasConcept C119857082 @default.
- W3176879439 hasConcept C127413603 @default.
- W3176879439 hasConcept C138885662 @default.
- W3176879439 hasConcept C150899416 @default.
- W3176879439 hasConcept C154945302 @default.
- W3176879439 hasConcept C177264268 @default.
- W3176879439 hasConcept C199360897 @default.
- W3176879439 hasConcept C201995342 @default.
- W3176879439 hasConcept C2776401178 @default.
- W3176879439 hasConcept C2780451532 @default.
- W3176879439 hasConcept C28006648 @default.
- W3176879439 hasConcept C41008148 @default.
- W3176879439 hasConcept C41895202 @default.
- W3176879439 hasConceptScore W3176879439C101104100 @default.
- W3176879439 hasConceptScore W3176879439C104409967 @default.
- W3176879439 hasConceptScore W3176879439C119857082 @default.
- W3176879439 hasConceptScore W3176879439C127413603 @default.
- W3176879439 hasConceptScore W3176879439C138885662 @default.
- W3176879439 hasConceptScore W3176879439C150899416 @default.
- W3176879439 hasConceptScore W3176879439C154945302 @default.
- W3176879439 hasConceptScore W3176879439C177264268 @default.
- W3176879439 hasConceptScore W3176879439C199360897 @default.
- W3176879439 hasConceptScore W3176879439C201995342 @default.
- W3176879439 hasConceptScore W3176879439C2776401178 @default.
- W3176879439 hasConceptScore W3176879439C2780451532 @default.
- W3176879439 hasConceptScore W3176879439C28006648 @default.
- W3176879439 hasConceptScore W3176879439C41008148 @default.
- W3176879439 hasConceptScore W3176879439C41895202 @default.
- W3176879439 hasIssue "18" @default.
- W3176879439 hasLocation W31768794391 @default.
- W3176879439 hasOpenAccess W3176879439 @default.
- W3176879439 hasPrimaryLocation W31768794391 @default.
- W3176879439 hasRelatedWork W2946016983 @default.
- W3176879439 hasRelatedWork W2960456850 @default.
- W3176879439 hasRelatedWork W2961085424 @default.
- W3176879439 hasRelatedWork W3021430260 @default.
- W3176879439 hasRelatedWork W3090469165 @default.
- W3176879439 hasRelatedWork W4281645081 @default.
- W3176879439 hasRelatedWork W4307205255 @default.
- W3176879439 hasRelatedWork W4308262314 @default.
- W3176879439 hasRelatedWork W4312200629 @default.
- W3176879439 hasRelatedWork W4382286161 @default.
- W3176879439 hasVolume "35" @default.
- W3176879439 isParatext "false" @default.
- W3176879439 isRetracted "false" @default.
- W3176879439 magId "3176879439" @default.
- W3176879439 workType "article" @default.