Matches in SemOpenAlex for { <https://semopenalex.org/work/W2725155646> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W2725155646 abstract "How can we learn a classifier that is fair for a protected or sensitive group, when we do not know if the input to the classifier belongs to the protected group? How can we train such a classifier when data on the protected group is difficult to attain? In many settings, finding out the sensitive input attribute can be prohibitively expensive even during model training, and sometimes impossible during model serving. For example, in recommender systems, if we want to predict if a user will click on a given recommendation, we often do not know many attributes of the user, e.g., race or age, and many attributes of the content are hard to determine, e.g., the language or topic. Thus, it is not feasible to use a different classifier calibrated based on knowledge of the sensitive attribute. Here, we use an adversarial training procedure to remove information about the sensitive attribute from the latent representation learned by a neural network. In particular, we study how the choice of data for the adversarial training effects the resulting fairness properties. We find two interesting results: a small amount of data is needed to train these adversarial models, and the data distribution empirically drives the adversary's notion of fairness." @default.
- W2725155646 created "2017-07-14" @default.
- W2725155646 creator A5015729655 @default.
- W2725155646 creator A5028125399 @default.
- W2725155646 creator A5033428202 @default.
- W2725155646 creator A5080988309 @default.
- W2725155646 date "2017-07-01" @default.
- W2725155646 modified "2023-09-25" @default.
- W2725155646 title "Data Decisions and Theoretical Implications when Adversarially Learning Fair Representations" @default.
- W2725155646 cites W1731081199 @default.
- W2725155646 cites W2146502635 @default.
- W2725155646 cites W2162670686 @default.
- W2725155646 cites W2247194987 @default.
- W2725155646 cites W2522104760 @default.
- W2725155646 cites W2604244242 @default.
- W2725155646 cites W2950018712 @default.
- W2725155646 cites W2950538796 @default.
- W2725155646 cites W2953127297 @default.
- W2725155646 cites W2962750142 @default.
- W2725155646 cites W3120740533 @default.
- W2725155646 hasPublicationYear "2017" @default.
- W2725155646 type Work @default.
- W2725155646 sameAs 2725155646 @default.
- W2725155646 citedByCount "153" @default.
- W2725155646 countsByYear W27251556462017 @default.
- W2725155646 countsByYear W27251556462018 @default.
- W2725155646 countsByYear W27251556462019 @default.
- W2725155646 countsByYear W27251556462020 @default.
- W2725155646 countsByYear W27251556462021 @default.
- W2725155646 countsByYear W27251556462022 @default.
- W2725155646 countsByYear W27251556462023 @default.
- W2725155646 crossrefType "posted-content" @default.
- W2725155646 hasAuthorship W2725155646A5015729655 @default.
- W2725155646 hasAuthorship W2725155646A5028125399 @default.
- W2725155646 hasAuthorship W2725155646A5033428202 @default.
- W2725155646 hasAuthorship W2725155646A5080988309 @default.
- W2725155646 hasConcept C119857082 @default.
- W2725155646 hasConcept C154945302 @default.
- W2725155646 hasConcept C37736160 @default.
- W2725155646 hasConcept C38652104 @default.
- W2725155646 hasConcept C41008148 @default.
- W2725155646 hasConcept C41065033 @default.
- W2725155646 hasConcept C50644808 @default.
- W2725155646 hasConcept C51632099 @default.
- W2725155646 hasConcept C95623464 @default.
- W2725155646 hasConceptScore W2725155646C119857082 @default.
- W2725155646 hasConceptScore W2725155646C154945302 @default.
- W2725155646 hasConceptScore W2725155646C37736160 @default.
- W2725155646 hasConceptScore W2725155646C38652104 @default.
- W2725155646 hasConceptScore W2725155646C41008148 @default.
- W2725155646 hasConceptScore W2725155646C41065033 @default.
- W2725155646 hasConceptScore W2725155646C50644808 @default.
- W2725155646 hasConceptScore W2725155646C51632099 @default.
- W2725155646 hasConceptScore W2725155646C95623464 @default.
- W2725155646 hasLocation W27251556461 @default.
- W2725155646 hasOpenAccess W2725155646 @default.
- W2725155646 hasPrimaryLocation W27251556461 @default.
- W2725155646 hasRelatedWork W1731081199 @default.
- W2725155646 hasRelatedWork W2014352947 @default.
- W2725155646 hasRelatedWork W2099471712 @default.
- W2725155646 hasRelatedWork W2100960835 @default.
- W2725155646 hasRelatedWork W2116666691 @default.
- W2725155646 hasRelatedWork W2116984840 @default.
- W2725155646 hasRelatedWork W2162670686 @default.
- W2725155646 hasRelatedWork W2483215953 @default.
- W2725155646 hasRelatedWork W2530395818 @default.
- W2725155646 hasRelatedWork W2540757487 @default.
- W2725155646 hasRelatedWork W2753845591 @default.
- W2725155646 hasRelatedWork W2788416960 @default.
- W2725155646 hasRelatedWork W2962750142 @default.
- W2725155646 hasRelatedWork W2963053914 @default.
- W2725155646 hasRelatedWork W2963116854 @default.
- W2725155646 hasRelatedWork W2963174898 @default.
- W2725155646 hasRelatedWork W2963178340 @default.
- W2725155646 hasRelatedWork W2963327716 @default.
- W2725155646 hasRelatedWork W2964031043 @default.
- W2725155646 hasRelatedWork W3023309920 @default.
- W2725155646 isParatext "false" @default.
- W2725155646 isRetracted "false" @default.
- W2725155646 magId "2725155646" @default.
- W2725155646 workType "article" @default.