Matches in SemOpenAlex for { <https://semopenalex.org/work/W4310921967> ?p ?o ?g. }
Showing items 1 to 62 of
62
with 100 items per page.
- W4310921967 abstract "Classifiers in supervised learning have various security and privacy issues, e.g., 1) data poisoning attacks, backdoor attacks, and adversarial examples on the security side as well as 2) inference attacks and the right to be forgotten for the training data on the privacy side. Various secure and privacy-preserving supervised learning algorithms with formal guarantees have been proposed to address these issues. However, they suffer from various limitations such as accuracy loss, small certified security guarantees, and/or inefficiency. Self-supervised learning is an emerging technique to pre-train encoders using unlabeled data. Given a pre-trained encoder as a feature extractor, supervised learning can train a simple yet accurate classifier using a small amount of labeled training data. In this work, we perform the first systematic, principled measurement study to understand whether and when a pre-trained encoder can address the limitations of secure or privacy-preserving supervised learning algorithms. Our key findings are that a pre-trained encoder substantially improves 1) both accuracy under no attacks and certified security guarantees against data poisoning and backdoor attacks of state-of-the-art secure learning algorithms (i.e., bagging and KNN), 2) certified security guarantees of randomized smoothing against adversarial examples without sacrificing its accuracy under no attacks, 3) accuracy of differentially private classifiers, and 4) accuracy and/or efficiency of exact machine unlearning." @default.
- W4310921967 created "2022-12-21" @default.
- W4310921967 creator A5009102659 @default.
- W4310921967 creator A5019571828 @default.
- W4310921967 creator A5056183335 @default.
- W4310921967 creator A5087464080 @default.
- W4310921967 date "2022-12-06" @default.
- W4310921967 modified "2023-10-16" @default.
- W4310921967 title "Pre-trained Encoders in Self-Supervised Learning Improve Secure and Privacy-preserving Supervised Learning" @default.
- W4310921967 doi "https://doi.org/10.48550/arxiv.2212.03334" @default.
- W4310921967 hasPublicationYear "2022" @default.
- W4310921967 type Work @default.
- W4310921967 citedByCount "0" @default.
- W4310921967 crossrefType "posted-content" @default.
- W4310921967 hasAuthorship W4310921967A5009102659 @default.
- W4310921967 hasAuthorship W4310921967A5019571828 @default.
- W4310921967 hasAuthorship W4310921967A5056183335 @default.
- W4310921967 hasAuthorship W4310921967A5087464080 @default.
- W4310921967 hasBestOaLocation W43109219671 @default.
- W4310921967 hasConcept C108583219 @default.
- W4310921967 hasConcept C111919701 @default.
- W4310921967 hasConcept C118505674 @default.
- W4310921967 hasConcept C119857082 @default.
- W4310921967 hasConcept C124101348 @default.
- W4310921967 hasConcept C136389625 @default.
- W4310921967 hasConcept C154945302 @default.
- W4310921967 hasConcept C2781045450 @default.
- W4310921967 hasConcept C38652104 @default.
- W4310921967 hasConcept C41008148 @default.
- W4310921967 hasConcept C50644808 @default.
- W4310921967 hasConcept C58973888 @default.
- W4310921967 hasConcept C95623464 @default.
- W4310921967 hasConceptScore W4310921967C108583219 @default.
- W4310921967 hasConceptScore W4310921967C111919701 @default.
- W4310921967 hasConceptScore W4310921967C118505674 @default.
- W4310921967 hasConceptScore W4310921967C119857082 @default.
- W4310921967 hasConceptScore W4310921967C124101348 @default.
- W4310921967 hasConceptScore W4310921967C136389625 @default.
- W4310921967 hasConceptScore W4310921967C154945302 @default.
- W4310921967 hasConceptScore W4310921967C2781045450 @default.
- W4310921967 hasConceptScore W4310921967C38652104 @default.
- W4310921967 hasConceptScore W4310921967C41008148 @default.
- W4310921967 hasConceptScore W4310921967C50644808 @default.
- W4310921967 hasConceptScore W4310921967C58973888 @default.
- W4310921967 hasConceptScore W4310921967C95623464 @default.
- W4310921967 hasLocation W43109219671 @default.
- W4310921967 hasLocation W43109219672 @default.
- W4310921967 hasOpenAccess W4310921967 @default.
- W4310921967 hasPrimaryLocation W43109219671 @default.
- W4310921967 hasRelatedWork W122912556 @default.
- W4310921967 hasRelatedWork W1586607209 @default.
- W4310921967 hasRelatedWork W2271357838 @default.
- W4310921967 hasRelatedWork W2348322200 @default.
- W4310921967 hasRelatedWork W2556866732 @default.
- W4310921967 hasRelatedWork W2621411691 @default.
- W4310921967 hasRelatedWork W3092753701 @default.
- W4310921967 hasRelatedWork W4281570223 @default.
- W4310921967 hasRelatedWork W4312414840 @default.
- W4310921967 hasRelatedWork W4385573555 @default.
- W4310921967 isParatext "false" @default.
- W4310921967 isRetracted "false" @default.
- W4310921967 workType "article" @default.