Matches in SemOpenAlex for { <https://semopenalex.org/work/W4300451640> ?p ?o ?g. }
Showing items 1 to 80 of
80
with 100 items per page.
- W4300451640 abstract "Machine Learning (ML) has emerged as a core technology to provide learning models to perform complex tasks. Boosted by Machine Learning as a Service (MLaaS), the number of applications relying on ML capabilities is ever increasing. However, ML models are the source of different privacy violations through passive or active attacks from different entities. In this paper, we present MixNN a proxy-based privacy-preserving system for federated learning to protect the privacy of participants against a curious or malicious aggregation server trying to infer sensitive information (i.e., membership and attribute inferences). MixNN receives the model updates from participants and mixes layers between participants before sending the mixed updates to the aggregation server. This mixing strategy drastically reduces privacy leaks without any trade-off with utility. Indeed, mixing the updates of the model has no impact on the result of the aggregation of the updates computed by the server. We report on an extensive evaluation of MixNN using several datasets and neural networks architectures to quantify privacy leakage through membership and attribute inference attacks as well the robustness of the protection. We show that MixNN significantly limits both the membership and attribute inferences compared to a baseline using model compression and noisy gradient (well known to damage the utility) while keeping the same level of utility as classic federated learning." @default.
- W4300451640 created "2022-10-03" @default.
- W4300451640 creator A5031657772 @default.
- W4300451640 creator A5038473707 @default.
- W4300451640 creator A5041298625 @default.
- W4300451640 creator A5072587955 @default.
- W4300451640 date "2022-11-07" @default.
- W4300451640 modified "2023-09-26" @default.
- W4300451640 title "MixNN" @default.
- W4300451640 cites W2051267297 @default.
- W4300451640 cites W2103647628 @default.
- W4300451640 cites W2145287260 @default.
- W4300451640 cites W2150248082 @default.
- W4300451640 cites W2211872228 @default.
- W4300451640 cites W2387306914 @default.
- W4300451640 cites W2473418344 @default.
- W4300451640 cites W2484015479 @default.
- W4300451640 cites W2568218703 @default.
- W4300451640 cites W2767079719 @default.
- W4300451640 cites W2793957616 @default.
- W4300451640 cites W2897830718 @default.
- W4300451640 cites W2930926105 @default.
- W4300451640 cites W2950656546 @default.
- W4300451640 cites W3096328345 @default.
- W4300451640 cites W3165040079 @default.
- W4300451640 cites W3176786489 @default.
- W4300451640 cites W4251569035 @default.
- W4300451640 doi "https://doi.org/10.1145/3528535.3565240" @default.
- W4300451640 hasPublicationYear "2022" @default.
- W4300451640 type Work @default.
- W4300451640 citedByCount "1" @default.
- W4300451640 countsByYear W43004516402023 @default.
- W4300451640 crossrefType "proceedings-article" @default.
- W4300451640 hasAuthorship W4300451640A5031657772 @default.
- W4300451640 hasAuthorship W4300451640A5038473707 @default.
- W4300451640 hasAuthorship W4300451640A5041298625 @default.
- W4300451640 hasAuthorship W4300451640A5072587955 @default.
- W4300451640 hasBestOaLocation W43004516402 @default.
- W4300451640 hasConcept C104317684 @default.
- W4300451640 hasConcept C119857082 @default.
- W4300451640 hasConcept C124101348 @default.
- W4300451640 hasConcept C154945302 @default.
- W4300451640 hasConcept C185592680 @default.
- W4300451640 hasConcept C2776214188 @default.
- W4300451640 hasConcept C2780148112 @default.
- W4300451640 hasConcept C2992525071 @default.
- W4300451640 hasConcept C38652104 @default.
- W4300451640 hasConcept C41008148 @default.
- W4300451640 hasConcept C55493867 @default.
- W4300451640 hasConcept C63479239 @default.
- W4300451640 hasConceptScore W4300451640C104317684 @default.
- W4300451640 hasConceptScore W4300451640C119857082 @default.
- W4300451640 hasConceptScore W4300451640C124101348 @default.
- W4300451640 hasConceptScore W4300451640C154945302 @default.
- W4300451640 hasConceptScore W4300451640C185592680 @default.
- W4300451640 hasConceptScore W4300451640C2776214188 @default.
- W4300451640 hasConceptScore W4300451640C2780148112 @default.
- W4300451640 hasConceptScore W4300451640C2992525071 @default.
- W4300451640 hasConceptScore W4300451640C38652104 @default.
- W4300451640 hasConceptScore W4300451640C41008148 @default.
- W4300451640 hasConceptScore W4300451640C55493867 @default.
- W4300451640 hasConceptScore W4300451640C63479239 @default.
- W4300451640 hasFunder F4320311618 @default.
- W4300451640 hasLocation W43004516401 @default.
- W4300451640 hasLocation W43004516402 @default.
- W4300451640 hasOpenAccess W4300451640 @default.
- W4300451640 hasPrimaryLocation W43004516401 @default.
- W4300451640 hasRelatedWork W2799803467 @default.
- W4300451640 hasRelatedWork W2961085424 @default.
- W4300451640 hasRelatedWork W2989877998 @default.
- W4300451640 hasRelatedWork W3046775127 @default.
- W4300451640 hasRelatedWork W4285260836 @default.
- W4300451640 hasRelatedWork W4286629047 @default.
- W4300451640 hasRelatedWork W4288019264 @default.
- W4300451640 hasRelatedWork W4306321456 @default.
- W4300451640 hasRelatedWork W4306674287 @default.
- W4300451640 hasRelatedWork W4224009465 @default.
- W4300451640 isParatext "false" @default.
- W4300451640 isRetracted "false" @default.
- W4300451640 workType "article" @default.