Matches in SemOpenAlex for { <https://semopenalex.org/work/W2051267297> ?p ?o ?g. }
Showing items 1 to 82 of
82
with 100 items per page.
- W2051267297 abstract "Machine-learning (ML) algorithms are increasingly utilized in privacy-sensitive applications such as predicting lifestyle choices, making medical diagnoses, and facial recognition. In a model inversion attack, recently introduced in a case study of linear classifiers in personalized medicine by Fredrikson et al., adversarial access to an ML model is abused to learn sensitive genomic information about individuals. Whether model inversion attacks apply to settings outside theirs, however, is unknown. We develop a new class of model inversion attack that exploits confidence values revealed along with predictions. Our new attacks are applicable in a variety of settings, and we explore two in depth: decision trees for lifestyle surveys as used on machine-learning-as-a-service systems and neural networks for facial recognition. In both cases confidence values are revealed to those with the ability to make prediction queries to models. We experimentally show attacks that are able to estimate whether a respondent in a lifestyle survey admitted to cheating on their significant other and, in the other context, show how to recover recognizable images of people's faces given only their name and access to the ML model. We also initiate experimental exploration of natural countermeasures, investigating a privacy-aware decision tree training algorithm that is a simple variant of CART learning, as well as revealing only rounded confidence values. The lesson that emerges is that one can avoid these kinds of MI attacks with negligible degradation to utility." @default.
- W2051267297 created "2016-06-24" @default.
- W2051267297 creator A5003774887 @default.
- W2051267297 creator A5057424614 @default.
- W2051267297 creator A5088826068 @default.
- W2051267297 date "2015-10-12" @default.
- W2051267297 modified "2023-10-18" @default.
- W2051267297 title "Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures" @default.
- W2051267297 cites W1515782956 @default.
- W2051267297 cites W2019735187 @default.
- W2051267297 cites W2040228409 @default.
- W2051267297 cites W2050164782 @default.
- W2051267297 cites W2075291208 @default.
- W2051267297 cites W2079115533 @default.
- W2051267297 cites W2095272373 @default.
- W2051267297 cites W2110868467 @default.
- W2051267297 cites W2120806354 @default.
- W2051267297 cites W2130325614 @default.
- W2051267297 cites W2135930857 @default.
- W2051267297 cites W2141481372 @default.
- W2051267297 cites W2141640676 @default.
- W2051267297 cites W2151298633 @default.
- W2051267297 cites W2160744452 @default.
- W2051267297 cites W2172013424 @default.
- W2051267297 cites W2296452361 @default.
- W2051267297 cites W2911978475 @default.
- W2051267297 cites W2963924367 @default.
- W2051267297 doi "https://doi.org/10.1145/2810103.2813677" @default.
- W2051267297 hasPublicationYear "2015" @default.
- W2051267297 type Work @default.
- W2051267297 sameAs 2051267297 @default.
- W2051267297 citedByCount "1519" @default.
- W2051267297 countsByYear W20512672972012 @default.
- W2051267297 countsByYear W20512672972016 @default.
- W2051267297 countsByYear W20512672972017 @default.
- W2051267297 countsByYear W20512672972018 @default.
- W2051267297 countsByYear W20512672972019 @default.
- W2051267297 countsByYear W20512672972020 @default.
- W2051267297 countsByYear W20512672972021 @default.
- W2051267297 countsByYear W20512672972022 @default.
- W2051267297 countsByYear W20512672972023 @default.
- W2051267297 crossrefType "proceedings-article" @default.
- W2051267297 hasAuthorship W2051267297A5003774887 @default.
- W2051267297 hasAuthorship W2051267297A5057424614 @default.
- W2051267297 hasAuthorship W2051267297A5088826068 @default.
- W2051267297 hasConcept C119857082 @default.
- W2051267297 hasConcept C142724271 @default.
- W2051267297 hasConcept C154945302 @default.
- W2051267297 hasConcept C165696696 @default.
- W2051267297 hasConcept C38652104 @default.
- W2051267297 hasConcept C41008148 @default.
- W2051267297 hasConcept C534262118 @default.
- W2051267297 hasConcept C71924100 @default.
- W2051267297 hasConcept C84525736 @default.
- W2051267297 hasConceptScore W2051267297C119857082 @default.
- W2051267297 hasConceptScore W2051267297C142724271 @default.
- W2051267297 hasConceptScore W2051267297C154945302 @default.
- W2051267297 hasConceptScore W2051267297C165696696 @default.
- W2051267297 hasConceptScore W2051267297C38652104 @default.
- W2051267297 hasConceptScore W2051267297C41008148 @default.
- W2051267297 hasConceptScore W2051267297C534262118 @default.
- W2051267297 hasConceptScore W2051267297C71924100 @default.
- W2051267297 hasConceptScore W2051267297C84525736 @default.
- W2051267297 hasFunder F4320306076 @default.
- W2051267297 hasFunder F4320332180 @default.
- W2051267297 hasLocation W20512672971 @default.
- W2051267297 hasOpenAccess W2051267297 @default.
- W2051267297 hasPrimaryLocation W20512672971 @default.
- W2051267297 hasRelatedWork W1470425429 @default.
- W2051267297 hasRelatedWork W3127425528 @default.
- W2051267297 hasRelatedWork W3185179407 @default.
- W2051267297 hasRelatedWork W4205478082 @default.
- W2051267297 hasRelatedWork W4281385048 @default.
- W2051267297 hasRelatedWork W4308191010 @default.
- W2051267297 hasRelatedWork W4313001487 @default.
- W2051267297 hasRelatedWork W4318350883 @default.
- W2051267297 hasRelatedWork W4328134586 @default.
- W2051267297 hasRelatedWork W4361795583 @default.
- W2051267297 isParatext "false" @default.
- W2051267297 isRetracted "false" @default.
- W2051267297 magId "2051267297" @default.
- W2051267297 workType "article" @default.