Matches in SemOpenAlex for { <https://semopenalex.org/work/W3092525263> ?p ?o ?g. }
Showing items 1 to 97 of
97
with 100 items per page.
- W3092525263 abstract "Machine learning as a service has emerged recently to relieve tensions between heavy deep learning tasks and increasing application demands. A deep learning service provider could help its clients to benefit from deep learning techniques at an affordable price instead of huge resource consumption. However, the service provider may have serious concerns about model privacy when a deep neural network model is published. Previous model publishing solutions mainly depend on additional artificial noise. By adding elaborated noises to parameters or gradients during the training phase, strong privacy guarantees like differential privacy could be achieved. However, this kind of approach cannot give guarantees on some other aspects, such as the quality of the disturbingly trained model and the convergence of the modified learning algorithm. In this paper, we propose an alternative private deep neural network model publishing solution, which caused no interference in the original training phase. We provide privacy, convergence and quality guarantees for the published model at the same time. Furthermore, our solution can achieve a smaller privacy budget when compared with artificial noise based training solutions proposed in previous works. Specifically, our solution gives an acceptable test accuracy with privacy budget ∊ = 1. Meanwhile, membership inference attack accuracy will be deceased from nearly 90% to around 60% across all classes." @default.
- W3092525263 created "2020-10-15" @default.
- W3092525263 creator A5017427585 @default.
- W3092525263 creator A5026169204 @default.
- W3092525263 creator A5044986337 @default.
- W3092525263 creator A5060739065 @default.
- W3092525263 creator A5086254999 @default.
- W3092525263 creator A5090706173 @default.
- W3092525263 date "2020-06-01" @default.
- W3092525263 modified "2023-09-24" @default.
- W3092525263 title "Private Deep Neural Network Models Publishing for Machine Learning as a Service" @default.
- W3092525263 cites W1489950266 @default.
- W3092525263 cites W1873763122 @default.
- W3092525263 cites W1993116423 @default.
- W3092525263 cites W2027595342 @default.
- W3092525263 cites W2051267297 @default.
- W3092525263 cites W2056784638 @default.
- W3092525263 cites W2112796928 @default.
- W3092525263 cites W2114766824 @default.
- W3092525263 cites W2473418344 @default.
- W3092525263 cites W2535690855 @default.
- W3092525263 cites W2591882872 @default.
- W3092525263 cites W2745269232 @default.
- W3092525263 cites W2767079719 @default.
- W3092525263 cites W2798657499 @default.
- W3092525263 cites W2806082141 @default.
- W3092525263 cites W2884943453 @default.
- W3092525263 cites W2897830718 @default.
- W3092525263 cites W2945335799 @default.
- W3092525263 cites W2963097630 @default.
- W3092525263 cites W2963228337 @default.
- W3092525263 cites W2963313259 @default.
- W3092525263 cites W2963456518 @default.
- W3092525263 cites W2963674932 @default.
- W3092525263 cites W2985580374 @default.
- W3092525263 cites W3102360395 @default.
- W3092525263 cites W3118608800 @default.
- W3092525263 doi "https://doi.org/10.1109/iwqos49365.2020.9212853" @default.
- W3092525263 hasPublicationYear "2020" @default.
- W3092525263 type Work @default.
- W3092525263 sameAs 3092525263 @default.
- W3092525263 citedByCount "2" @default.
- W3092525263 countsByYear W30925252632021 @default.
- W3092525263 crossrefType "proceedings-article" @default.
- W3092525263 hasAuthorship W3092525263A5017427585 @default.
- W3092525263 hasAuthorship W3092525263A5026169204 @default.
- W3092525263 hasAuthorship W3092525263A5044986337 @default.
- W3092525263 hasAuthorship W3092525263A5060739065 @default.
- W3092525263 hasAuthorship W3092525263A5086254999 @default.
- W3092525263 hasAuthorship W3092525263A5090706173 @default.
- W3092525263 hasConcept C108583219 @default.
- W3092525263 hasConcept C11413529 @default.
- W3092525263 hasConcept C116537 @default.
- W3092525263 hasConcept C119857082 @default.
- W3092525263 hasConcept C136264566 @default.
- W3092525263 hasConcept C154945302 @default.
- W3092525263 hasConcept C162324750 @default.
- W3092525263 hasConcept C23130292 @default.
- W3092525263 hasConcept C2777303404 @default.
- W3092525263 hasConcept C2780378061 @default.
- W3092525263 hasConcept C31258907 @default.
- W3092525263 hasConcept C41008148 @default.
- W3092525263 hasConcept C50522688 @default.
- W3092525263 hasConcept C50644808 @default.
- W3092525263 hasConcept C5119721 @default.
- W3092525263 hasConceptScore W3092525263C108583219 @default.
- W3092525263 hasConceptScore W3092525263C11413529 @default.
- W3092525263 hasConceptScore W3092525263C116537 @default.
- W3092525263 hasConceptScore W3092525263C119857082 @default.
- W3092525263 hasConceptScore W3092525263C136264566 @default.
- W3092525263 hasConceptScore W3092525263C154945302 @default.
- W3092525263 hasConceptScore W3092525263C162324750 @default.
- W3092525263 hasConceptScore W3092525263C23130292 @default.
- W3092525263 hasConceptScore W3092525263C2777303404 @default.
- W3092525263 hasConceptScore W3092525263C2780378061 @default.
- W3092525263 hasConceptScore W3092525263C31258907 @default.
- W3092525263 hasConceptScore W3092525263C41008148 @default.
- W3092525263 hasConceptScore W3092525263C50522688 @default.
- W3092525263 hasConceptScore W3092525263C50644808 @default.
- W3092525263 hasConceptScore W3092525263C5119721 @default.
- W3092525263 hasLocation W30925252631 @default.
- W3092525263 hasOpenAccess W3092525263 @default.
- W3092525263 hasPrimaryLocation W30925252631 @default.
- W3092525263 hasRelatedWork W10786582 @default.
- W3092525263 hasRelatedWork W13678974 @default.
- W3092525263 hasRelatedWork W256534 @default.
- W3092525263 hasRelatedWork W6680660 @default.
- W3092525263 hasRelatedWork W8021486 @default.
- W3092525263 hasRelatedWork W9190101 @default.
- W3092525263 hasRelatedWork W9333608 @default.
- W3092525263 hasRelatedWork W1512708 @default.
- W3092525263 hasRelatedWork W4825420 @default.
- W3092525263 hasRelatedWork W7759631 @default.
- W3092525263 isParatext "false" @default.
- W3092525263 isRetracted "false" @default.
- W3092525263 magId "3092525263" @default.
- W3092525263 workType "article" @default.