Deep Asymmetric Multi-Task Feature Learning

Cited 15 time in webofscience Cited 0 time in scopus
  • Hit : 100
  • Download : 0
DC FieldValueLanguage
dc.contributor.authorLee, Hae Beomko
dc.contributor.authorYang, Eunhoko
dc.contributor.authorHwang, Sung Juko
dc.date.accessioned2022-01-14T06:57:32Z-
dc.date.available2022-01-14T06:57:32Z-
dc.date.created2021-12-01-
dc.date.issued2018-07-12-
dc.identifier.citation35th International Conference on Machine Learning, ICML 2018, pp.4609 - 4619-
dc.identifier.issn2640-3498-
dc.identifier.urihttp://hdl.handle.net/10203/291845-
dc.description.abstractWe propose Deep Asymmetric Multitask Feature Learning (Deep-AMTFL) which can learn deep representations shared across multiple tasks while effectively preventing negative transfer that may happen in the feature sharing process. Specifically, we introduce an asymmetric autoencoder term that allows reliable predictors for the easy tasks to have high contribution to the feature learning while suppressing the influences of unreliable predictors for more difficult tasks. This allows the learning of less noisy representations, and enables unreliable predictors to exploit knowledge from the reliable predictors via the shared latent features. Such asymmetric knowledge transfer through shared features is also more scalable and efficient than inter-task asymmetric transfer. We validate our Deep-AMTFL model on multiple benchmark datasets for multitask learning and image classification, on which it significantly outperforms existing symmetric and asymmetric multitask learning models, by effectively preventing negative transfer in deep feature learning.-
dc.languageEnglish-
dc.publisherInternational Machine Learning Society (IMLS)-
dc.titleDeep Asymmetric Multi-Task Feature Learning-
dc.typeConference-
dc.identifier.wosid000683379203008-
dc.identifier.scopusid2-s2.0-85057240594-
dc.type.rimsCONF-
dc.citation.beginningpage4609-
dc.citation.endingpage4619-
dc.citation.publicationname35th International Conference on Machine Learning, ICML 2018-
dc.identifier.conferencecountrySW-
dc.identifier.conferencelocationStockholm-
dc.contributor.localauthorYang, Eunho-
dc.contributor.localauthorHwang, Sung Ju-
Appears in Collection
RIMS Conference Papers
Files in This Item
There are no files associated with this item.
This item is cited by other documents in WoS
⊙ Detail Information in WoSⓡ Click to see webofscience_button
⊙ Cited 15 items in WoS Click to see citing articles in records_button

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0