Matches in SemOpenAlex for { <https://semopenalex.org/work/W2977527445> ?p ?o ?g. }
- W2977527445 abstract "Federated Learning (FL) systems are gaining popularity as a solution to training Machine Learning (ML) models from large-scale user data collected on personal devices (e.g., smartphones) without their raw data leaving the device. At the core of FL is a network of anonymous user devices sharing training information (model parameter updates) computed locally on personal data. However, the type and degree to which user-specific information is encoded in the model updates is poorly understood. In this paper, we identify model updates encode subtle variations in which users capture and generate data. The variations provide a strong statistical signal, allowing an adversary to effectively deanonymize participating devices using a limited set of auxiliary data. We analyze resulting deanonymization attacks on diverse tasks on real-world (anonymized) user-generated data across a range of closed- and open-world scenarios. We study various strategies to mitigate the risks of deanonymization. As random perturbation methods do not offer convincing operating points, we propose data-augmentation strategies which introduces adversarial biases in device data and thereby, offer substantial protection against deanonymization threats with little effect on utility." @default.
- W2977527445 created "2019-10-10" @default.
- W2977527445 creator A5003887059 @default.
- W2977527445 creator A5022802322 @default.
- W2977527445 creator A5025851635 @default.
- W2977527445 creator A5051534545 @default.
- W2977527445 creator A5087547624 @default.
- W2977527445 date "2018-05-15" @default.
- W2977527445 modified "2023-10-05" @default.
- W2977527445 title "Gradient-Leaks: Understanding and Controlling Deanonymization in Federated Learning" @default.
- W2977527445 cites W133884053 @default.
- W2977527445 cites W1528350527 @default.
- W2977527445 cites W1600806036 @default.
- W2977527445 cites W1762484328 @default.
- W2977527445 cites W1767117126 @default.
- W2977527445 cites W1849277567 @default.
- W2977527445 cites W1861492603 @default.
- W2977527445 cites W1892323599 @default.
- W2977527445 cites W1895577753 @default.
- W2977527445 cites W1903029394 @default.
- W2977527445 cites W196349407 @default.
- W2977527445 cites W1968059489 @default.
- W2977527445 cites W1968697355 @default.
- W2977527445 cites W1975937116 @default.
- W2977527445 cites W1978209044 @default.
- W2977527445 cites W2026910242 @default.
- W2977527445 cites W2031489346 @default.
- W2977527445 cites W2051267297 @default.
- W2977527445 cites W2053637704 @default.
- W2977527445 cites W2088492763 @default.
- W2977527445 cites W2090659861 @default.
- W2977527445 cites W2097117768 @default.
- W2977527445 cites W2101234009 @default.
- W2977527445 cites W2102086994 @default.
- W2977527445 cites W2103018059 @default.
- W2977527445 cites W2106053110 @default.
- W2977527445 cites W2108598243 @default.
- W2977527445 cites W2109426455 @default.
- W2977527445 cites W2110302976 @default.
- W2977527445 cites W2113839561 @default.
- W2977527445 cites W2117154949 @default.
- W2977527445 cites W2119404697 @default.
- W2977527445 cites W2123154536 @default.
- W2977527445 cites W2127589108 @default.
- W2977527445 cites W2130556178 @default.
- W2977527445 cites W2132339004 @default.
- W2977527445 cites W2138302120 @default.
- W2977527445 cites W2157364932 @default.
- W2977527445 cites W2163605009 @default.
- W2977527445 cites W2165424625 @default.
- W2977527445 cites W2168231600 @default.
- W2977527445 cites W2171590421 @default.
- W2977527445 cites W2187089797 @default.
- W2977527445 cites W2194775991 @default.
- W2977527445 cites W2296452361 @default.
- W2977527445 cites W2340382781 @default.
- W2977527445 cites W2341535507 @default.
- W2977527445 cites W2435473771 @default.
- W2977527445 cites W2481726180 @default.
- W2977527445 cites W2524365899 @default.
- W2977527445 cites W2530417694 @default.
- W2977527445 cites W2535690855 @default.
- W2977527445 cites W2535838896 @default.
- W2977527445 cites W2557728737 @default.
- W2977527445 cites W2560545764 @default.
- W2977527445 cites W2570343428 @default.
- W2977527445 cites W2591882872 @default.
- W2977527445 cites W2599765304 @default.
- W2977527445 cites W2612445135 @default.
- W2977527445 cites W2743388417 @default.
- W2977527445 cites W2752337926 @default.
- W2977527445 cites W2755682530 @default.
- W2977527445 cites W2767079719 @default.
- W2977527445 cites W2767368215 @default.
- W2977527445 cites W2773907782 @default.
- W2977527445 cites W2777914285 @default.
- W2977527445 cites W2783000019 @default.
- W2977527445 cites W2784621220 @default.
- W2977527445 cites W2799803467 @default.
- W2977527445 cites W2888940765 @default.
- W2977527445 cites W2890697686 @default.
- W2977527445 cites W2890991187 @default.
- W2977527445 cites W2897830718 @default.
- W2977527445 cites W2897959453 @default.
- W2977527445 cites W2900120080 @default.
- W2977527445 cites W2903389359 @default.
- W2977527445 cites W2904190483 @default.
- W2977527445 cites W2905455284 @default.
- W2977527445 cites W2911978475 @default.
- W2977527445 cites W2913340405 @default.
- W2977527445 cites W2924921501 @default.
- W2977527445 cites W2949338291 @default.
- W2977527445 cites W2950193633 @default.
- W2977527445 cites W2950239670 @default.
- W2977527445 cites W2950602864 @default.
- W2977527445 cites W2950925835 @default.
- W2977527445 cites W2950959228 @default.
- W2977527445 cites W2951055356 @default.
- W2977527445 cites W2951762694 @default.
- W2977527445 cites W2952927437 @default.