Matches in SemOpenAlex for { <https://semopenalex.org/work/W4226296584> ?p ?o ?g. }
Showing items 1 to 61 of
61
with 100 items per page.
- W4226296584 abstract "Machine learning models are trained to minimize the mean loss for a single metric, and thus typically do not consider fairness and robustness. Neglecting such metrics in training can make these models prone to fairness violations when training data are imbalanced or test distributions differ. This work introduces Fairness Optimized Reweighting via Meta-Learning (FORML), a training algorithm that balances fairness and robustness with accuracy by jointly learning training sample weights and neural network parameters. The approach increases model fairness by learning to balance the contributions from both over- and under-represented sub-groups through dynamic reweighting of the data learned from a user-specified held-out set representative of the distribution under which fairness is desired. FORML improves equality of opportunity fairness criteria on image classification tasks, reduces bias of corrupted labels, and facilitates building more fair datasets via data condensation. These improvements are achieved without pre-processing data or post-processing model outputs, without learning an additional weighting function, without changing model architecture, and while maintaining accuracy on the original predictive metric." @default.
- W4226296584 created "2022-05-05" @default.
- W4226296584 creator A5019401621 @default.
- W4226296584 creator A5059839283 @default.
- W4226296584 creator A5071541814 @default.
- W4226296584 date "2022-02-03" @default.
- W4226296584 modified "2023-09-24" @default.
- W4226296584 title "FORML: Learning to Reweight Data for Fairness" @default.
- W4226296584 doi "https://doi.org/10.48550/arxiv.2202.01719" @default.
- W4226296584 hasPublicationYear "2022" @default.
- W4226296584 type Work @default.
- W4226296584 citedByCount "0" @default.
- W4226296584 crossrefType "posted-content" @default.
- W4226296584 hasAuthorship W4226296584A5019401621 @default.
- W4226296584 hasAuthorship W4226296584A5059839283 @default.
- W4226296584 hasAuthorship W4226296584A5071541814 @default.
- W4226296584 hasBestOaLocation W42262965841 @default.
- W4226296584 hasConcept C104317684 @default.
- W4226296584 hasConcept C119857082 @default.
- W4226296584 hasConcept C126838900 @default.
- W4226296584 hasConcept C154945302 @default.
- W4226296584 hasConcept C162324750 @default.
- W4226296584 hasConcept C176217482 @default.
- W4226296584 hasConcept C183115368 @default.
- W4226296584 hasConcept C185592680 @default.
- W4226296584 hasConcept C21547014 @default.
- W4226296584 hasConcept C41008148 @default.
- W4226296584 hasConcept C50644808 @default.
- W4226296584 hasConcept C55493867 @default.
- W4226296584 hasConcept C63479239 @default.
- W4226296584 hasConcept C71924100 @default.
- W4226296584 hasConceptScore W4226296584C104317684 @default.
- W4226296584 hasConceptScore W4226296584C119857082 @default.
- W4226296584 hasConceptScore W4226296584C126838900 @default.
- W4226296584 hasConceptScore W4226296584C154945302 @default.
- W4226296584 hasConceptScore W4226296584C162324750 @default.
- W4226296584 hasConceptScore W4226296584C176217482 @default.
- W4226296584 hasConceptScore W4226296584C183115368 @default.
- W4226296584 hasConceptScore W4226296584C185592680 @default.
- W4226296584 hasConceptScore W4226296584C21547014 @default.
- W4226296584 hasConceptScore W4226296584C41008148 @default.
- W4226296584 hasConceptScore W4226296584C50644808 @default.
- W4226296584 hasConceptScore W4226296584C55493867 @default.
- W4226296584 hasConceptScore W4226296584C63479239 @default.
- W4226296584 hasConceptScore W4226296584C71924100 @default.
- W4226296584 hasLocation W42262965841 @default.
- W4226296584 hasOpenAccess W4226296584 @default.
- W4226296584 hasPrimaryLocation W42262965841 @default.
- W4226296584 hasRelatedWork W1556035623 @default.
- W4226296584 hasRelatedWork W2066966161 @default.
- W4226296584 hasRelatedWork W2111353840 @default.
- W4226296584 hasRelatedWork W2369301223 @default.
- W4226296584 hasRelatedWork W2961085424 @default.
- W4226296584 hasRelatedWork W3084651595 @default.
- W4226296584 hasRelatedWork W4286629047 @default.
- W4226296584 hasRelatedWork W4306674287 @default.
- W4226296584 hasRelatedWork W1629725936 @default.
- W4226296584 hasRelatedWork W4224009465 @default.
- W4226296584 isParatext "false" @default.
- W4226296584 isRetracted "false" @default.
- W4226296584 workType "article" @default.