FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313117512> ?p ?o ?g. }

• W4313117512 endingPage "432" @default.
• W4313117512 startingPage "414" @default.
• W4313117512 abstract "Existing pruning techniques preserve deep neural networks’ overall ability to make correct predictions but could also amplify hidden biases during the compression process. We propose a novel pruning method, Fairness-aware GRAdient Pruning mEthod (FairGRAPE), that minimizes the disproportionate impacts of pruning on different sub-groups. Our method calculates the per-group importance of each model weight and selects a subset of weights that maintain the relative between-group total importance in pruning. The proposed method then prunes network edges with small importance values and repeats the procedure by updating importance values. We demonstrate the effectiveness of our method on four different datasets, FairFace, UTKFace, CelebA, and ImageNet, for the tasks of face attribute classification where our method reduces the disparity in performance degradation by up to 90% compared to the state-of-the-art pruning algorithms. Our method is substantially more effective in a setting with a high pruning rate (99%). The code and dataset used in the experiments are available at https://github.com/Bernardo1998/FairGRAPE" @default.
• W4313117512 created "2023-01-06" @default.
• W4313117512 creator A5000684736 @default.
• W4313117512 creator A5010677168 @default.
• W4313117512 creator A5011230999 @default.
• W4313117512 date "2022-01-01" @default.
• W4313117512 modified "2023-10-10" @default.
• W4313117512 title "FairGRAPE: Fairness-Aware GRAdient Pruning mEthod for Face Attribute Classification" @default.
• W4313117512 cites W1834627138 @default.
• W4313117512 cites W2108598243 @default.
• W4313117512 cites W2194775991 @default.
• W4313117512 cites W2592232824 @default.
• W4313117512 cites W2883070812 @default.
• W4313117512 cites W2890680318 @default.
• W4313117512 cites W2914036220 @default.
• W4313117512 cites W2948264761 @default.
• W4313117512 cites W2962749380 @default.
• W4313117512 cites W2962787423 @default.
• W4313117512 cites W2962851801 @default.
• W4313117512 cites W2962883549 @default.
• W4313117512 cites W2963145730 @default.
• W4313117512 cites W2963363373 @default.
• W4313117512 cites W2963942586 @default.
• W4313117512 cites W2965174861 @default.
• W4313117512 cites W2965862774 @default.
• W4313117512 cites W2982232682 @default.
• W4313117512 cites W2982358316 @default.
• W4313117512 cites W2990751682 @default.
• W4313117512 cites W3006210081 @default.
• W4313117512 cites W3034700241 @default.
• W4313117512 cites W3035447285 @default.
• W4313117512 cites W3037771725 @default.
• W4313117512 cites W3092661284 @default.
• W4313117512 cites W3120331640 @default.
• W4313117512 cites W3120485916 @default.
• W4313117512 cites W3133226796 @default.
• W4313117512 cites W3140832216 @default.
• W4313117512 cites W3168398407 @default.
• W4313117512 cites W3172872502 @default.
• W4313117512 cites W3175129878 @default.
• W4313117512 cites W3202521030 @default.
• W4313117512 cites W3216710733 @default.
• W4313117512 cites W3217183037 @default.
• W4313117512 cites W4283028305 @default.
• W4313117512 cites W4288083803 @default.
• W4313117512 cites W4312991156 @default.
• W4313117512 doi "https://doi.org/10.1007/978-3-031-19778-9_24" @default.
• W4313117512 hasPublicationYear "2022" @default.
• W4313117512 type Work @default.
• W4313117512 citedByCount "4" @default.
• W4313117512 countsByYear W43131175122023 @default.
• W4313117512 crossrefType "book-chapter" @default.
• W4313117512 hasAuthorship W4313117512A5000684736 @default.
• W4313117512 hasAuthorship W4313117512A5010677168 @default.
• W4313117512 hasAuthorship W4313117512A5011230999 @default.
• W4313117512 hasBestOaLocation W43131175122 @default.
• W4313117512 hasConcept C108010975 @default.
• W4313117512 hasConcept C111919701 @default.
• W4313117512 hasConcept C119857082 @default.
• W4313117512 hasConcept C144024400 @default.
• W4313117512 hasConcept C153180895 @default.
• W4313117512 hasConcept C154945302 @default.
• W4313117512 hasConcept C177264268 @default.
• W4313117512 hasConcept C199360897 @default.
• W4313117512 hasConcept C2776760102 @default.
• W4313117512 hasConcept C2779304628 @default.
• W4313117512 hasConcept C2984842247 @default.
• W4313117512 hasConcept C36289849 @default.
• W4313117512 hasConcept C41008148 @default.
• W4313117512 hasConcept C50644808 @default.
• W4313117512 hasConcept C6557445 @default.
• W4313117512 hasConcept C86803240 @default.
• W4313117512 hasConcept C98045186 @default.
• W4313117512 hasConceptScore W4313117512C108010975 @default.
• W4313117512 hasConceptScore W4313117512C111919701 @default.
• W4313117512 hasConceptScore W4313117512C119857082 @default.
• W4313117512 hasConceptScore W4313117512C144024400 @default.
• W4313117512 hasConceptScore W4313117512C153180895 @default.
• W4313117512 hasConceptScore W4313117512C154945302 @default.
• W4313117512 hasConceptScore W4313117512C177264268 @default.
• W4313117512 hasConceptScore W4313117512C199360897 @default.
• W4313117512 hasConceptScore W4313117512C2776760102 @default.
• W4313117512 hasConceptScore W4313117512C2779304628 @default.
• W4313117512 hasConceptScore W4313117512C2984842247 @default.
• W4313117512 hasConceptScore W4313117512C36289849 @default.
• W4313117512 hasConceptScore W4313117512C41008148 @default.
• W4313117512 hasConceptScore W4313117512C50644808 @default.
• W4313117512 hasConceptScore W4313117512C6557445 @default.
• W4313117512 hasConceptScore W4313117512C86803240 @default.
• W4313117512 hasConceptScore W4313117512C98045186 @default.
• W4313117512 hasLocation W43131175121 @default.
• W4313117512 hasLocation W43131175122 @default.
• W4313117512 hasOpenAccess W4313117512 @default.
• W4313117512 hasPrimaryLocation W43131175121 @default.
• W4313117512 hasRelatedWork W1490470283 @default.
• W4313117512 hasRelatedWork W1775397219 @default.
• W4313117512 hasRelatedWork W2347601237 @default.
• W4313117512 hasRelatedWork W2383164569 @default.

• next