Matches in SemOpenAlex for { <https://semopenalex.org/work/W2922917409> ?p ?o ?g. }
Showing items 1 to 85 of
85
with 100 items per page.
- W2922917409 abstract "In computer vision, virtually every state-of-the-art deep learning system is trained with data augmentation. In text classification, however, data augmentation is less widely practiced because it must be performed before training and risks introducing label noise. We augment the IMDB movie reviews dataset with examples generated by two families of techniques: random token perturbations introduced by Wei and Zou [2019] and backtranslation -- translating to a second language then back to English. In low resource environments, backtranslation generates significant improvement on top of the state of-the-art ULMFit model. A ULMFit model pretrained on wikitext103 and then fine-tuned on only 50 IMDB examples and 500 synthetic examples generated by backtranslation achieves 80.6% accuracy, an 8.1% improvement over the augmentation-free baseline with only 9 minutes of additional training time. Random token perturbations do not yield any improvements but incur equivalent computational cost. The benefits of training with backtranslated examples decreases with the size of the available training data. On the full dataset, neither augmentation technique improves upon ULMFit's state of the art performance. We address this by using backtranslations as a form of test time augmentation as well as ensembling ULMFit with other models, and achieve small improvements." @default.
- W2922917409 created "2019-04-01" @default.
- W2922917409 creator A5042449490 @default.
- W2922917409 date "2019-03-21" @default.
- W2922917409 modified "2023-09-27" @default.
- W2922917409 title "Low Resource Text Classification with ULMFit and Backtranslation." @default.
- W2922917409 cites W2113459411 @default.
- W2922917409 cites W2394658926 @default.
- W2922917409 cites W2525332836 @default.
- W2922917409 cites W2743945814 @default.
- W2922917409 cites W2889326796 @default.
- W2922917409 cites W2963026768 @default.
- W2922917409 cites W2963083752 @default.
- W2922917409 cites W2963216553 @default.
- W2922917409 cites W2963341956 @default.
- W2922917409 cites W2971296908 @default.
- W2922917409 hasPublicationYear "2019" @default.
- W2922917409 type Work @default.
- W2922917409 sameAs 2922917409 @default.
- W2922917409 citedByCount "13" @default.
- W2922917409 countsByYear W29229174092019 @default.
- W2922917409 countsByYear W29229174092020 @default.
- W2922917409 countsByYear W29229174092021 @default.
- W2922917409 countsByYear W29229174092023 @default.
- W2922917409 crossrefType "posted-content" @default.
- W2922917409 hasAuthorship W2922917409A5042449490 @default.
- W2922917409 hasConcept C111368507 @default.
- W2922917409 hasConcept C11413529 @default.
- W2922917409 hasConcept C119857082 @default.
- W2922917409 hasConcept C12725497 @default.
- W2922917409 hasConcept C127313418 @default.
- W2922917409 hasConcept C137293760 @default.
- W2922917409 hasConcept C154945302 @default.
- W2922917409 hasConcept C169258074 @default.
- W2922917409 hasConcept C206345919 @default.
- W2922917409 hasConcept C28490314 @default.
- W2922917409 hasConcept C31258907 @default.
- W2922917409 hasConcept C38652104 @default.
- W2922917409 hasConcept C41008148 @default.
- W2922917409 hasConcept C48103436 @default.
- W2922917409 hasConcept C48145219 @default.
- W2922917409 hasConcept C51632099 @default.
- W2922917409 hasConceptScore W2922917409C111368507 @default.
- W2922917409 hasConceptScore W2922917409C11413529 @default.
- W2922917409 hasConceptScore W2922917409C119857082 @default.
- W2922917409 hasConceptScore W2922917409C12725497 @default.
- W2922917409 hasConceptScore W2922917409C127313418 @default.
- W2922917409 hasConceptScore W2922917409C137293760 @default.
- W2922917409 hasConceptScore W2922917409C154945302 @default.
- W2922917409 hasConceptScore W2922917409C169258074 @default.
- W2922917409 hasConceptScore W2922917409C206345919 @default.
- W2922917409 hasConceptScore W2922917409C28490314 @default.
- W2922917409 hasConceptScore W2922917409C31258907 @default.
- W2922917409 hasConceptScore W2922917409C38652104 @default.
- W2922917409 hasConceptScore W2922917409C41008148 @default.
- W2922917409 hasConceptScore W2922917409C48103436 @default.
- W2922917409 hasConceptScore W2922917409C48145219 @default.
- W2922917409 hasConceptScore W2922917409C51632099 @default.
- W2922917409 hasLocation W29229174091 @default.
- W2922917409 hasOpenAccess W2922917409 @default.
- W2922917409 hasPrimaryLocation W29229174091 @default.
- W2922917409 hasRelatedWork W1832693441 @default.
- W2922917409 hasRelatedWork W2070246124 @default.
- W2922917409 hasRelatedWork W2250539671 @default.
- W2922917409 hasRelatedWork W2251939518 @default.
- W2922917409 hasRelatedWork W2889326796 @default.
- W2922917409 hasRelatedWork W2954428350 @default.
- W2922917409 hasRelatedWork W2963026768 @default.
- W2922917409 hasRelatedWork W2963216553 @default.
- W2922917409 hasRelatedWork W2963341956 @default.
- W2922917409 hasRelatedWork W2963403868 @default.
- W2922917409 hasRelatedWork W2963545917 @default.
- W2922917409 hasRelatedWork W2965373594 @default.
- W2922917409 hasRelatedWork W2971252690 @default.
- W2922917409 hasRelatedWork W2971296908 @default.
- W2922917409 hasRelatedWork W2978916561 @default.
- W2922917409 hasRelatedWork W2980370789 @default.
- W2922917409 hasRelatedWork W3096660185 @default.
- W2922917409 hasRelatedWork W3143887455 @default.
- W2922917409 hasRelatedWork W3169141681 @default.
- W2922917409 hasRelatedWork W3209906346 @default.
- W2922917409 isParatext "false" @default.
- W2922917409 isRetracted "false" @default.
- W2922917409 magId "2922917409" @default.
- W2922917409 workType "article" @default.