FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4312182206> ?p ?o ?g. }

• W4312182206 endingPage "102819" @default.
• W4312182206 startingPage "102819" @default.
• W4312182206 abstract "We propose a novel approach for achieving differential privacy for neural network training models through compression and memorization of gradients. The compression technique, which makes gradient vectors sparse, reduces the sensitivity so that differential privacy can be achieved with less noise; whereas the memorization technique, which remembers unused gradient parts, keeps track of the descent direction and thereby maintains the accuracy of the proposed algorithm. Our differentially private algorithm, called dp-memSGD for short, converges mathematically at the same rate of 1/T as standard stochastic gradient descent (SGD) algorithm, where T is the number of training iterations. Experimentally, we demonstrate that dp-memSGD converges with reasonable privacy losses on many benchmark datasets." @default.
• W4312182206 created "2023-01-04" @default.
• W4312182206 creator A5034778794 @default.
• W4312182206 creator A5068171354 @default.
• W4312182206 date "2023-02-01" @default.
• W4312182206 modified "2023-09-25" @default.
• W4312182206 title "Differentially private stochastic gradient descent via compression and memorization" @default.
• W4312182206 cites W1873763122 @default.
• W4312182206 cites W1985511977 @default.
• W4312182206 cites W1992926795 @default.
• W4312182206 cites W2053637704 @default.
• W4312182206 cites W2091825929 @default.
• W4312182206 cites W2194775991 @default.
• W4312182206 cites W2295292576 @default.
• W4312182206 cites W2481704219 @default.
• W4312182206 cites W2520442116 @default.
• W4312182206 cites W2535690855 @default.
• W4312182206 cites W2595094219 @default.
• W4312182206 cites W2758152646 @default.
• W4312182206 cites W2781091734 @default.
• W4312182206 cites W2884943453 @default.
• W4312182206 cites W2913608349 @default.
• W4312182206 cites W2955023110 @default.
• W4312182206 cites W2963433607 @default.
• W4312182206 cites W2963470657 @default.
• W4312182206 cites W2963952467 @default.
• W4312182206 cites W2963999993 @default.
• W4312182206 cites W2964756884 @default.
• W4312182206 cites W3024346083 @default.
• W4312182206 cites W3035174002 @default.
• W4312182206 cites W3047250731 @default.
• W4312182206 cites W3048427923 @default.
• W4312182206 cites W3102360395 @default.
• W4312182206 cites W3110543402 @default.
• W4312182206 cites W3127040002 @default.
• W4312182206 cites W3128083813 @default.
• W4312182206 cites W3175645569 @default.
• W4312182206 cites W3191743800 @default.
• W4312182206 cites W4225264920 @default.
• W4312182206 cites W4280494975 @default.
• W4312182206 cites W4289262024 @default.
• W4312182206 cites W4293783406 @default.
• W4312182206 doi "https://doi.org/10.1016/j.sysarc.2022.102819" @default.
• W4312182206 hasPublicationYear "2023" @default.
• W4312182206 type Work @default.
• W4312182206 citedByCount "3" @default.
• W4312182206 countsByYear W43121822062023 @default.
• W4312182206 crossrefType "journal-article" @default.
• W4312182206 hasAuthorship W4312182206A5034778794 @default.
• W4312182206 hasAuthorship W4312182206A5068171354 @default.
• W4312182206 hasConcept C11413529 @default.
• W4312182206 hasConcept C13280743 @default.
• W4312182206 hasConcept C145420912 @default.
• W4312182206 hasConcept C153258448 @default.
• W4312182206 hasConcept C154945302 @default.
• W4312182206 hasConcept C159985019 @default.
• W4312182206 hasConcept C180016635 @default.
• W4312182206 hasConcept C185798385 @default.
• W4312182206 hasConcept C192562407 @default.
• W4312182206 hasConcept C205649164 @default.
• W4312182206 hasConcept C206688291 @default.
• W4312182206 hasConcept C23130292 @default.
• W4312182206 hasConcept C30038468 @default.
• W4312182206 hasConcept C33923547 @default.
• W4312182206 hasConcept C41008148 @default.
• W4312182206 hasConcept C50644808 @default.
• W4312182206 hasConceptScore W4312182206C11413529 @default.
• W4312182206 hasConceptScore W4312182206C13280743 @default.
• W4312182206 hasConceptScore W4312182206C145420912 @default.
• W4312182206 hasConceptScore W4312182206C153258448 @default.
• W4312182206 hasConceptScore W4312182206C154945302 @default.
• W4312182206 hasConceptScore W4312182206C159985019 @default.
• W4312182206 hasConceptScore W4312182206C180016635 @default.
• W4312182206 hasConceptScore W4312182206C185798385 @default.
• W4312182206 hasConceptScore W4312182206C192562407 @default.
• W4312182206 hasConceptScore W4312182206C205649164 @default.
• W4312182206 hasConceptScore W4312182206C206688291 @default.
• W4312182206 hasConceptScore W4312182206C23130292 @default.
• W4312182206 hasConceptScore W4312182206C30038468 @default.
• W4312182206 hasConceptScore W4312182206C33923547 @default.
• W4312182206 hasConceptScore W4312182206C41008148 @default.
• W4312182206 hasConceptScore W4312182206C50644808 @default.
• W4312182206 hasFunder F4320334789 @default.
• W4312182206 hasFunder F4320338075 @default.
• W4312182206 hasLocation W43121822061 @default.
• W4312182206 hasOpenAccess W4312182206 @default.
• W4312182206 hasPrimaryLocation W43121822061 @default.
• W4312182206 hasRelatedWork W1985511977 @default.
• W4312182206 hasRelatedWork W2552083804 @default.
• W4312182206 hasRelatedWork W2998988480 @default.
• W4312182206 hasRelatedWork W3094098295 @default.
• W4312182206 hasRelatedWork W3107554447 @default.
• W4312182206 hasRelatedWork W3128458668 @default.
• W4312182206 hasRelatedWork W3170244987 @default.
• W4312182206 hasRelatedWork W4285709917 @default.
• W4312182206 hasRelatedWork W4309134251 @default.
• W4312182206 hasRelatedWork W4312182206 @default.
• W4312182206 hasVolume "135" @default.

• next