Matches in SemOpenAlex for { <https://semopenalex.org/work/W2897830718> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W2897830718 abstract "With the growing adoption of machine learning, sharing of learned models is becoming popular. However, in addition to the prediction properties the model producer aims to share, there is also a risk that the model consumer can infer other properties of the training data the model producer did not intend to share. In this paper, we focus on the inference of global properties of the training data, such as the environment in which the data was produced, or the fraction of the data that comes from a certain class, as applied to white-box Fully Connected Neural Networks (FCNNs). Because of their complexity and inscrutability, FCNNs have a particularly high risk of leaking unexpected information about their training sets; at the same time, this complexity makes extracting this information challenging. We develop techniques that reduce this complexity by noting that FCNNs are invariant under permutation of nodes in each layer. We develop our techniques using representations that capture this invariance and simplify the information extraction task. We evaluate our techniques on several synthetic and standard benchmark datasets and show that they are very effective at inferring various data properties. We also perform two case studies to demonstrate the impact of our attack. In the first case study we show that a classifier that recognizes smiling faces also leaks information about the relative attractiveness of the individuals in its training set. In the second case study we show that a classifier that recognizes Bitcoin mining from performance counters also leaks information about whether the classifier was trained on logs from machines that were patched for the Meltdown and Spectre attacks." @default.
- W2897830718 created "2018-10-26" @default.
- W2897830718 creator A5006956415 @default.
- W2897830718 creator A5015195367 @default.
- W2897830718 creator A5024942018 @default.
- W2897830718 creator A5031954035 @default.
- W2897830718 creator A5066827571 @default.
- W2897830718 date "2018-10-15" @default.
- W2897830718 modified "2023-10-12" @default.
- W2897830718 title "Property Inference Attacks on Fully Connected Neural Networks using Permutation Invariant Representations" @default.
- W2897830718 cites W2051267297 @default.
- W2897830718 cites W2071289869 @default.
- W2897830718 cites W2080592089 @default.
- W2897830718 cites W2132862423 @default.
- W2897830718 cites W2145287260 @default.
- W2897830718 cites W2243397390 @default.
- W2897830718 cites W2473418344 @default.
- W2897830718 cites W2522718524 @default.
- W2897830718 cites W2535690855 @default.
- W2897830718 cites W2603766943 @default.
- W2897830718 cites W2757528734 @default.
- W2897830718 cites W2763172007 @default.
- W2897830718 cites W2962835266 @default.
- W2897830718 cites W3007346474 @default.
- W2897830718 cites W3099206234 @default.
- W2897830718 doi "https://doi.org/10.1145/3243734.3243834" @default.
- W2897830718 hasPublicationYear "2018" @default.
- W2897830718 type Work @default.
- W2897830718 sameAs 2897830718 @default.
- W2897830718 citedByCount "226" @default.
- W2897830718 countsByYear W28978307182018 @default.
- W2897830718 countsByYear W28978307182019 @default.
- W2897830718 countsByYear W28978307182020 @default.
- W2897830718 countsByYear W28978307182021 @default.
- W2897830718 countsByYear W28978307182022 @default.
- W2897830718 countsByYear W28978307182023 @default.
- W2897830718 crossrefType "proceedings-article" @default.
- W2897830718 hasAuthorship W2897830718A5006956415 @default.
- W2897830718 hasAuthorship W2897830718A5015195367 @default.
- W2897830718 hasAuthorship W2897830718A5024942018 @default.
- W2897830718 hasAuthorship W2897830718A5031954035 @default.
- W2897830718 hasAuthorship W2897830718A5066827571 @default.
- W2897830718 hasBestOaLocation W28978307181 @default.
- W2897830718 hasConcept C111472728 @default.
- W2897830718 hasConcept C121332964 @default.
- W2897830718 hasConcept C138885662 @default.
- W2897830718 hasConcept C154945302 @default.
- W2897830718 hasConcept C189950617 @default.
- W2897830718 hasConcept C190470478 @default.
- W2897830718 hasConcept C21308566 @default.
- W2897830718 hasConcept C24890656 @default.
- W2897830718 hasConcept C2776214188 @default.
- W2897830718 hasConcept C33923547 @default.
- W2897830718 hasConcept C37914503 @default.
- W2897830718 hasConcept C41008148 @default.
- W2897830718 hasConcept C50644808 @default.
- W2897830718 hasConcept C80444323 @default.
- W2897830718 hasConceptScore W2897830718C111472728 @default.
- W2897830718 hasConceptScore W2897830718C121332964 @default.
- W2897830718 hasConceptScore W2897830718C138885662 @default.
- W2897830718 hasConceptScore W2897830718C154945302 @default.
- W2897830718 hasConceptScore W2897830718C189950617 @default.
- W2897830718 hasConceptScore W2897830718C190470478 @default.
- W2897830718 hasConceptScore W2897830718C21308566 @default.
- W2897830718 hasConceptScore W2897830718C24890656 @default.
- W2897830718 hasConceptScore W2897830718C2776214188 @default.
- W2897830718 hasConceptScore W2897830718C33923547 @default.
- W2897830718 hasConceptScore W2897830718C37914503 @default.
- W2897830718 hasConceptScore W2897830718C41008148 @default.
- W2897830718 hasConceptScore W2897830718C50644808 @default.
- W2897830718 hasConceptScore W2897830718C80444323 @default.
- W2897830718 hasFunder F4320337674 @default.
- W2897830718 hasLocation W28978307181 @default.
- W2897830718 hasOpenAccess W2897830718 @default.
- W2897830718 hasPrimaryLocation W28978307181 @default.
- W2897830718 hasRelatedWork W1671699713 @default.
- W2897830718 hasRelatedWork W2367950322 @default.
- W2897830718 hasRelatedWork W2386387936 @default.
- W2897830718 hasRelatedWork W2394233781 @default.
- W2897830718 hasRelatedWork W2950054037 @default.
- W2897830718 hasRelatedWork W3107474891 @default.
- W2897830718 hasRelatedWork W3121749535 @default.
- W2897830718 hasRelatedWork W3127983536 @default.
- W2897830718 hasRelatedWork W3163360311 @default.
- W2897830718 hasRelatedWork W4225504845 @default.
- W2897830718 isParatext "false" @default.
- W2897830718 isRetracted "false" @default.
- W2897830718 magId "2897830718" @default.
- W2897830718 workType "article" @default.