Matches in SemOpenAlex for { <https://semopenalex.org/work/W2945180690> ?p ?o ?g. }
- W2945180690 abstract "Transferring knowledge from one neural network to another has been shown to be helpful for learning tasks with few training examples. Prevailing fine-tuning methods could potentially contaminate pre-trained features by comparably high energy random noise. This noise is mainly delivered from a careless replacement of task-specific parameters. We analyze theoretically such knowledge contamination for classification tasks and propose a practical and easy to apply method to trap and minimize the contaminant. In our approach, the entropy of the output estimates gets maximized initially and the first back-propagated error is stalled at the output of the last layer. Our proposed method not only outperforms the traditional fine-tuning, but also significantly speeds up the convergence of the learner. It is robust to randomness and independent of the choice of architecture. Overall, our experiments show that the power of transfer learning has been substantially underestimated so far." @default.
- W2945180690 created "2019-05-29" @default.
- W2945180690 creator A5002021405 @default.
- W2945180690 creator A5035036000 @default.
- W2945180690 creator A5036056813 @default.
- W2945180690 creator A5042893723 @default.
- W2945180690 creator A5065059896 @default.
- W2945180690 date "2019-05-25" @default.
- W2945180690 modified "2023-09-27" @default.
- W2945180690 title "Efficient Neural Task Adaptation by Maximum Entropy Initialization." @default.
- W2945180690 cites W1533861849 @default.
- W2945180690 cites W1677182931 @default.
- W2945180690 cites W1686810756 @default.
- W2945180690 cites W2062118960 @default.
- W2945180690 cites W2102605133 @default.
- W2945180690 cites W2160921898 @default.
- W2945180690 cites W2166049352 @default.
- W2945180690 cites W2183341477 @default.
- W2945180690 cites W2194775991 @default.
- W2945180690 cites W2203224402 @default.
- W2945180690 cites W2473930607 @default.
- W2945180690 cites W2910114055 @default.
- W2945180690 cites W2949117887 @default.
- W2945180690 cites W2953360861 @default.
- W2945180690 cites W2963446712 @default.
- W2945180690 cites W3118608800 @default.
- W2945180690 cites W3146803896 @default.
- W2945180690 hasPublicationYear "2019" @default.
- W2945180690 type Work @default.
- W2945180690 sameAs 2945180690 @default.
- W2945180690 citedByCount "2" @default.
- W2945180690 countsByYear W29451806902019 @default.
- W2945180690 countsByYear W29451806902020 @default.
- W2945180690 crossrefType "posted-content" @default.
- W2945180690 hasAuthorship W2945180690A5002021405 @default.
- W2945180690 hasAuthorship W2945180690A5035036000 @default.
- W2945180690 hasAuthorship W2945180690A5036056813 @default.
- W2945180690 hasAuthorship W2945180690A5042893723 @default.
- W2945180690 hasAuthorship W2945180690A5065059896 @default.
- W2945180690 hasConcept C105795698 @default.
- W2945180690 hasConcept C106301342 @default.
- W2945180690 hasConcept C114466953 @default.
- W2945180690 hasConcept C119857082 @default.
- W2945180690 hasConcept C120665830 @default.
- W2945180690 hasConcept C121332964 @default.
- W2945180690 hasConcept C125112378 @default.
- W2945180690 hasConcept C127413603 @default.
- W2945180690 hasConcept C139807058 @default.
- W2945180690 hasConcept C150899416 @default.
- W2945180690 hasConcept C154945302 @default.
- W2945180690 hasConcept C199360897 @default.
- W2945180690 hasConcept C201995342 @default.
- W2945180690 hasConcept C2780451532 @default.
- W2945180690 hasConcept C33923547 @default.
- W2945180690 hasConcept C41008148 @default.
- W2945180690 hasConcept C50644808 @default.
- W2945180690 hasConcept C62520636 @default.
- W2945180690 hasConcept C9679016 @default.
- W2945180690 hasConceptScore W2945180690C105795698 @default.
- W2945180690 hasConceptScore W2945180690C106301342 @default.
- W2945180690 hasConceptScore W2945180690C114466953 @default.
- W2945180690 hasConceptScore W2945180690C119857082 @default.
- W2945180690 hasConceptScore W2945180690C120665830 @default.
- W2945180690 hasConceptScore W2945180690C121332964 @default.
- W2945180690 hasConceptScore W2945180690C125112378 @default.
- W2945180690 hasConceptScore W2945180690C127413603 @default.
- W2945180690 hasConceptScore W2945180690C139807058 @default.
- W2945180690 hasConceptScore W2945180690C150899416 @default.
- W2945180690 hasConceptScore W2945180690C154945302 @default.
- W2945180690 hasConceptScore W2945180690C199360897 @default.
- W2945180690 hasConceptScore W2945180690C201995342 @default.
- W2945180690 hasConceptScore W2945180690C2780451532 @default.
- W2945180690 hasConceptScore W2945180690C33923547 @default.
- W2945180690 hasConceptScore W2945180690C41008148 @default.
- W2945180690 hasConceptScore W2945180690C50644808 @default.
- W2945180690 hasConceptScore W2945180690C62520636 @default.
- W2945180690 hasConceptScore W2945180690C9679016 @default.
- W2945180690 hasLocation W29451806901 @default.
- W2945180690 hasOpenAccess W2945180690 @default.
- W2945180690 hasPrimaryLocation W29451806901 @default.
- W2945180690 hasRelatedWork W2188295774 @default.
- W2945180690 hasRelatedWork W2626439280 @default.
- W2945180690 hasRelatedWork W2726756416 @default.
- W2945180690 hasRelatedWork W2743693791 @default.
- W2945180690 hasRelatedWork W2752860440 @default.
- W2945180690 hasRelatedWork W2752971446 @default.
- W2945180690 hasRelatedWork W2768666611 @default.
- W2945180690 hasRelatedWork W2781596748 @default.
- W2945180690 hasRelatedWork W2802636049 @default.
- W2945180690 hasRelatedWork W2901104906 @default.
- W2945180690 hasRelatedWork W2914317573 @default.
- W2945180690 hasRelatedWork W2916965824 @default.
- W2945180690 hasRelatedWork W2946742785 @default.
- W2945180690 hasRelatedWork W2951388117 @default.
- W2945180690 hasRelatedWork W2963939958 @default.
- W2945180690 hasRelatedWork W2990596457 @default.
- W2945180690 hasRelatedWork W3003494558 @default.
- W2945180690 hasRelatedWork W3130602232 @default.
- W2945180690 hasRelatedWork W3164079106 @default.
- W2945180690 hasRelatedWork W3208436298 @default.