FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3017048210> ?p ?o ?g. }

• W3017048210 abstract "Computer vision algorithms, e.g. for face recognition, favour groups of individuals that are better represented in the training data. This happens because of the generalization that classifiers have to make. It is simpler to fit the majority groups as this fit is more important to overall error. We propose to create a balanced training dataset, consisting of the original dataset plus new data points in which the group memberships are intervened, minorities become majorities and vice versa. We show that current generative adversarial networks are a powerful tool for learning these data points, called contrastive examples. We experiment with the equalized odds bias measure on tabular data as well as image data (CelebA and Diversity in Faces datasets). Contrastive examples allow us to expose correlations between group membership and other seemingly neutral features. Whenever a causal graph is available, we can put those contrastive examples in the perspective of counterfactuals." @default.
• W3017048210 created "2020-04-24" @default.
• W3017048210 creator A5006827927 @default.
• W3017048210 creator A5020758501 @default.
• W3017048210 creator A5021434291 @default.
• W3017048210 creator A5029105520 @default.
• W3017048210 date "2020-04-14" @default.
• W3017048210 modified "2023-09-27" @default.
• W3017048210 title "Contrastive Examples for Addressing the Tyranny of the Majority" @default.
• W3017048210 cites W1523116602 @default.
• W3017048210 cites W1525651898 @default.
• W3017048210 cites W1731081199 @default.
• W3017048210 cites W1834627138 @default.
• W3017048210 cites W1961345416 @default.
• W3017048210 cites W2099471712 @default.
• W3017048210 cites W2116984840 @default.
• W3017048210 cites W2147947791 @default.
• W3017048210 cites W2150291618 @default.
• W3017048210 cites W2162670686 @default.
• W3017048210 cites W2194775991 @default.
• W3017048210 cites W2212660284 @default.
• W3017048210 cites W2214409633 @default.
• W3017048210 cites W2522957395 @default.
• W3017048210 cites W2530395818 @default.
• W3017048210 cites W2540757487 @default.
• W3017048210 cites W2592232824 @default.
• W3017048210 cites W2725155646 @default.
• W3017048210 cites W2739748921 @default.
• W3017048210 cites W2767437859 @default.
• W3017048210 cites W2767657961 @default.
• W3017048210 cites W2786122898 @default.
• W3017048210 cites W2788416960 @default.
• W3017048210 cites W2808309499 @default.
• W3017048210 cites W2810873347 @default.
• W3017048210 cites W2890680318 @default.
• W3017048210 cites W2897978524 @default.
• W3017048210 cites W2914454064 @default.
• W3017048210 cites W2937229771 @default.
• W3017048210 cites W2950772592 @default.
• W3017048210 cites W2962793481 @default.
• W3017048210 cites W2962858109 @default.
• W3017048210 cites W2962879692 @default.
• W3017048210 cites W2962922665 @default.
• W3017048210 cites W2962977061 @default.
• W3017048210 cites W2963016445 @default.
• W3017048210 cites W2963116854 @default.
• W3017048210 cites W2963327716 @default.
• W3017048210 cites W2963446520 @default.
• W3017048210 cites W2963709863 @default.
• W3017048210 cites W2963767194 @default.
• W3017048210 cites W2963803533 @default.
• W3017048210 cites W2963818033 @default.
• W3017048210 cites W2963839617 @default.
• W3017048210 cites W2964031043 @default.
• W3017048210 cites W2982232682 @default.
• W3017048210 cites W3037302481 @default.
• W3017048210 cites W3041475347 @default.
• W3017048210 cites W3122175177 @default.
• W3017048210 cites W3123374861 @default.
• W3017048210 hasPublicationYear "2020" @default.
• W3017048210 type Work @default.
• W3017048210 sameAs 3017048210 @default.
• W3017048210 citedByCount "3" @default.
• W3017048210 countsByYear W30170482102020 @default.
• W3017048210 countsByYear W30170482102021 @default.
• W3017048210 crossrefType "posted-content" @default.
• W3017048210 hasAuthorship W3017048210A5006827927 @default.
• W3017048210 hasAuthorship W3017048210A5020758501 @default.
• W3017048210 hasAuthorship W3017048210A5021434291 @default.
• W3017048210 hasAuthorship W3017048210A5029105520 @default.
• W3017048210 hasConcept C119857082 @default.
• W3017048210 hasConcept C12713177 @default.
• W3017048210 hasConcept C134306372 @default.
• W3017048210 hasConcept C143095724 @default.
• W3017048210 hasConcept C151956035 @default.
• W3017048210 hasConcept C153180895 @default.
• W3017048210 hasConcept C154945302 @default.
• W3017048210 hasConcept C177148314 @default.
• W3017048210 hasConcept C178790620 @default.
• W3017048210 hasConcept C185592680 @default.
• W3017048210 hasConcept C204321447 @default.
• W3017048210 hasConcept C2781311116 @default.
• W3017048210 hasConcept C33923547 @default.
• W3017048210 hasConcept C37736160 @default.
• W3017048210 hasConcept C39890363 @default.
• W3017048210 hasConcept C41008148 @default.
• W3017048210 hasConceptScore W3017048210C119857082 @default.
• W3017048210 hasConceptScore W3017048210C12713177 @default.
• W3017048210 hasConceptScore W3017048210C134306372 @default.
• W3017048210 hasConceptScore W3017048210C143095724 @default.
• W3017048210 hasConceptScore W3017048210C151956035 @default.
• W3017048210 hasConceptScore W3017048210C153180895 @default.
• W3017048210 hasConceptScore W3017048210C154945302 @default.
• W3017048210 hasConceptScore W3017048210C177148314 @default.
• W3017048210 hasConceptScore W3017048210C178790620 @default.
• W3017048210 hasConceptScore W3017048210C185592680 @default.
• W3017048210 hasConceptScore W3017048210C204321447 @default.
• W3017048210 hasConceptScore W3017048210C2781311116 @default.
• W3017048210 hasConceptScore W3017048210C33923547 @default.
• W3017048210 hasConceptScore W3017048210C37736160 @default.

• next