Matches in SemOpenAlex for { <https://semopenalex.org/work/W3189925607> ?p ?o ?g. }
- W3189925607 abstract "Dataset bias is a critical challenge in machine learning since it often leads to a negative impact on a model due to the unintended decision rules captured by spurious correlations. Although existing works often handle this issue based on human supervision, the availability of the proper annotations is impractical and even unrealistic. To better tackle the limitation, we propose a simple but effective unsupervised debiasing technique. Specifically, we first identify pseudo-attributes based on the results from clustering performed in the feature embedding space even without an explicit bias attribute supervision. Then, we employ a novel cluster-wise reweighting scheme to learn debiased representation; the proposed method prevents minority groups from being discounted for minimizing the overall loss, which is desirable for worst-case generalization. The extensive experiments demonstrate the outstanding performance of our approach on multiple standard benchmarks, even achieving the competitive accuracy to the supervised counterpart." @default.
- W3189925607 created "2021-08-16" @default.
- W3189925607 creator A5006594639 @default.
- W3189925607 creator A5044267273 @default.
- W3189925607 creator A5068063341 @default.
- W3189925607 date "2021-08-06" @default.
- W3189925607 modified "2023-09-24" @default.
- W3189925607 title "Unsupervised Learning of Debiased Representations with Pseudo-Attributes" @default.
- W3189925607 cites W1484551447 @default.
- W3189925607 cites W1797268635 @default.
- W3189925607 cites W1834627138 @default.
- W3189925607 cites W2031342017 @default.
- W3189925607 cites W2162670686 @default.
- W3189925607 cites W2531499355 @default.
- W3189925607 cites W2561715562 @default.
- W3189925607 cites W2732026016 @default.
- W3189925607 cites W2763549966 @default.
- W3189925607 cites W2781132081 @default.
- W3189925607 cites W2788557041 @default.
- W3189925607 cites W2794583223 @default.
- W3189925607 cites W2895243423 @default.
- W3189925607 cites W2897978524 @default.
- W3189925607 cites W2902617128 @default.
- W3189925607 cites W2907520152 @default.
- W3189925607 cites W2912592113 @default.
- W3189925607 cites W2946193510 @default.
- W3189925607 cites W2958360136 @default.
- W3189925607 cites W2963263347 @default.
- W3189925607 cites W2963349562 @default.
- W3189925607 cites W2963803533 @default.
- W3189925607 cites W2970017794 @default.
- W3189925607 cites W2970019270 @default.
- W3189925607 cites W2982232682 @default.
- W3189925607 cites W2982358316 @default.
- W3189925607 cites W2990751682 @default.
- W3189925607 cites W2991497298 @default.
- W3189925607 cites W3016970897 @default.
- W3189925607 cites W3035037113 @default.
- W3189925607 cites W3037363477 @default.
- W3189925607 cites W3038239361 @default.
- W3189925607 cites W3038745855 @default.
- W3189925607 cites W3044799724 @default.
- W3189925607 cites W3046022822 @default.
- W3189925607 cites W3100350094 @default.
- W3189925607 cites W3109831211 @default.
- W3189925607 cites W3113580345 @default.
- W3189925607 cites W3114796327 @default.
- W3189925607 doi "https://doi.org/10.48550/arxiv.2108.02943" @default.
- W3189925607 hasPublicationYear "2021" @default.
- W3189925607 type Work @default.
- W3189925607 sameAs 3189925607 @default.
- W3189925607 citedByCount "1" @default.
- W3189925607 countsByYear W31899256072021 @default.
- W3189925607 crossrefType "posted-content" @default.
- W3189925607 hasAuthorship W3189925607A5006594639 @default.
- W3189925607 hasAuthorship W3189925607A5044267273 @default.
- W3189925607 hasAuthorship W3189925607A5068063341 @default.
- W3189925607 hasBestOaLocation W31899256071 @default.
- W3189925607 hasConcept C111472728 @default.
- W3189925607 hasConcept C119857082 @default.
- W3189925607 hasConcept C124101348 @default.
- W3189925607 hasConcept C134306372 @default.
- W3189925607 hasConcept C138885662 @default.
- W3189925607 hasConcept C154945302 @default.
- W3189925607 hasConcept C15744967 @default.
- W3189925607 hasConcept C177148314 @default.
- W3189925607 hasConcept C17744445 @default.
- W3189925607 hasConcept C188147891 @default.
- W3189925607 hasConcept C199539241 @default.
- W3189925607 hasConcept C2776359362 @default.
- W3189925607 hasConcept C2776401178 @default.
- W3189925607 hasConcept C2779458634 @default.
- W3189925607 hasConcept C2780586882 @default.
- W3189925607 hasConcept C33923547 @default.
- W3189925607 hasConcept C41008148 @default.
- W3189925607 hasConcept C41608201 @default.
- W3189925607 hasConcept C41895202 @default.
- W3189925607 hasConcept C59404180 @default.
- W3189925607 hasConcept C73555534 @default.
- W3189925607 hasConcept C77618280 @default.
- W3189925607 hasConcept C8038995 @default.
- W3189925607 hasConcept C83665646 @default.
- W3189925607 hasConcept C94625758 @default.
- W3189925607 hasConcept C97256817 @default.
- W3189925607 hasConceptScore W3189925607C111472728 @default.
- W3189925607 hasConceptScore W3189925607C119857082 @default.
- W3189925607 hasConceptScore W3189925607C124101348 @default.
- W3189925607 hasConceptScore W3189925607C134306372 @default.
- W3189925607 hasConceptScore W3189925607C138885662 @default.
- W3189925607 hasConceptScore W3189925607C154945302 @default.
- W3189925607 hasConceptScore W3189925607C15744967 @default.
- W3189925607 hasConceptScore W3189925607C177148314 @default.
- W3189925607 hasConceptScore W3189925607C17744445 @default.
- W3189925607 hasConceptScore W3189925607C188147891 @default.
- W3189925607 hasConceptScore W3189925607C199539241 @default.
- W3189925607 hasConceptScore W3189925607C2776359362 @default.
- W3189925607 hasConceptScore W3189925607C2776401178 @default.
- W3189925607 hasConceptScore W3189925607C2779458634 @default.
- W3189925607 hasConceptScore W3189925607C2780586882 @default.
- W3189925607 hasConceptScore W3189925607C33923547 @default.