Matches in SemOpenAlex for { <https://semopenalex.org/work/W4315706034> ?p ?o ?g. }
Showing items 1 to 59 of
59
with 100 items per page.
- W4315706034 endingPage "84" @default.
- W4315706034 startingPage "68" @default.
- W4315706034 abstract "We study the privacy implications of training recurrent neural networks (RNNs) with sensitive training datasets. Considering membership inference attacks (MIAs)—which aim to infer whether or not specific data records have been used in training a given machine learning model—we provide empirical evidence that a neural network's architecture impacts its vulnerability to MIAs. In particular, we demonstrate that RNNs are subject to a higher attack accuracy than feed-forward neural network (FFNN) counterparts. Additionally, we study the effectiveness of two prominent mitigation methods for preempting MIAs, namely weight regularization and differential privacy. For the former, we empirically demonstrate that RNNs may only benefit from weight regularization marginally as opposed to FFNNs. For the latter, we find that enforcing differential privacy through either of the following two methods leads to a less favorable privacy-utility trade-off in RNNs than alternative FFNNs: (i) adding Gaussian noise to the gradients calculated during training as a part of the so-called extsc{DP-SGD} algorithm and (ii) adding Gaussian noise to the trainable parameters as a part of a post-training mechanism that we propose. As a result, RNNs can also be less amenable to mitigation methods, bringing us to the conclusion that the privacy risks pertaining to the recurrent architecture are higher than the feed-forward counterparts." @default.
- W4315706034 created "2023-01-12" @default.
- W4315706034 creator A5042227524 @default.
- W4315706034 creator A5047471973 @default.
- W4315706034 creator A5071606862 @default.
- W4315706034 date "2023-01-01" @default.
- W4315706034 modified "2023-10-05" @default.
- W4315706034 title "On the Privacy Risks of Deploying Recurrent Neural Networks in Machine Learning Models" @default.
- W4315706034 doi "https://doi.org/10.56553/popets-2023-0005" @default.
- W4315706034 hasPublicationYear "2023" @default.
- W4315706034 type Work @default.
- W4315706034 citedByCount "0" @default.
- W4315706034 crossrefType "journal-article" @default.
- W4315706034 hasAuthorship W4315706034A5042227524 @default.
- W4315706034 hasAuthorship W4315706034A5047471973 @default.
- W4315706034 hasAuthorship W4315706034A5071606862 @default.
- W4315706034 hasBestOaLocation W43157060341 @default.
- W4315706034 hasConcept C119857082 @default.
- W4315706034 hasConcept C124101348 @default.
- W4315706034 hasConcept C147168706 @default.
- W4315706034 hasConcept C154945302 @default.
- W4315706034 hasConcept C23130292 @default.
- W4315706034 hasConcept C2776135515 @default.
- W4315706034 hasConcept C2776214188 @default.
- W4315706034 hasConcept C2984842247 @default.
- W4315706034 hasConcept C41008148 @default.
- W4315706034 hasConcept C50644808 @default.
- W4315706034 hasConceptScore W4315706034C119857082 @default.
- W4315706034 hasConceptScore W4315706034C124101348 @default.
- W4315706034 hasConceptScore W4315706034C147168706 @default.
- W4315706034 hasConceptScore W4315706034C154945302 @default.
- W4315706034 hasConceptScore W4315706034C23130292 @default.
- W4315706034 hasConceptScore W4315706034C2776135515 @default.
- W4315706034 hasConceptScore W4315706034C2776214188 @default.
- W4315706034 hasConceptScore W4315706034C2984842247 @default.
- W4315706034 hasConceptScore W4315706034C41008148 @default.
- W4315706034 hasConceptScore W4315706034C50644808 @default.
- W4315706034 hasIssue "1" @default.
- W4315706034 hasLocation W43157060341 @default.
- W4315706034 hasLocation W43157060342 @default.
- W4315706034 hasLocation W43157060343 @default.
- W4315706034 hasOpenAccess W4315706034 @default.
- W4315706034 hasPrimaryLocation W43157060341 @default.
- W4315706034 hasRelatedWork W1591801644 @default.
- W4315706034 hasRelatedWork W2765920929 @default.
- W4315706034 hasRelatedWork W2952270003 @default.
- W4315706034 hasRelatedWork W3082444838 @default.
- W4315706034 hasRelatedWork W3154070777 @default.
- W4315706034 hasRelatedWork W4281386417 @default.
- W4315706034 hasRelatedWork W4311992262 @default.
- W4315706034 hasRelatedWork W4327831767 @default.
- W4315706034 hasRelatedWork W4372260407 @default.
- W4315706034 hasRelatedWork W4382201660 @default.
- W4315706034 hasVolume "2023" @default.
- W4315706034 isParatext "false" @default.
- W4315706034 isRetracted "false" @default.
- W4315706034 workType "article" @default.