Dynamically balancing class losses in imbalanced deep learning
Dynamically balancing class losses in imbalanced deep learning
About this item
Full title
Author / Creator
Publisher
Stevenage: John Wiley & Sons, Inc
Journal title
Language
English
Formats
Publication information
Publisher
Stevenage: John Wiley & Sons, Inc
Subjects
More information
Scope and Contents
Contents
Imbalanced datasets commonly exist in real world and greatly challenge the performances of deep neural models. The authors discover that the traditional balance strategies in the existing imbalanced learning methods emphasize/suppress the importance of minority/majority class in a fixed way. Even the minority class is fully represented, the minorit...
Alternative Titles
Full title
Dynamically balancing class losses in imbalanced deep learning
Authors, Artists and Contributors
Author / Creator
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_doaj_primary_oai_doaj_org_article_28a3901c98354a5184fc42f1acc6b3b1
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_doaj_primary_oai_doaj_org_article_28a3901c98354a5184fc42f1acc6b3b1
Other Identifiers
ISSN
0013-5194
E-ISSN
1350-911X
DOI
10.1049/ell2.12408
