Matches in SemOpenAlex for { <https://semopenalex.org/work/W4226523637> ?p ?o ?g. }
Showing items 1 to 55 of
55
with 100 items per page.
- W4226523637 abstract "Split learning is a popular technique used for vertical federated learning (VFL), where the goal is to jointly train a model on the private input and label data held by two parties. This technique uses a split-model, trained end-to-end, by exchanging the intermediate representations (IR) of the inputs and gradients of the IR between the two parties. We propose ExPLoit - a label-leakage attack that allows an adversarial input-owner to extract the private labels of the label-owner during split-learning. ExPLoit frames the attack as a supervised learning problem by using a novel loss function that combines gradient-matching and several regularization terms developed using key properties of the dataset and models. Our evaluations show that ExPLoit can uncover the private labels with near-perfect accuracy of up to 99.96%. Our findings underscore the need for better training techniques for VFL." @default.
- W4226523637 created "2022-05-05" @default.
- W4226523637 creator A5060352379 @default.
- W4226523637 creator A5082772077 @default.
- W4226523637 date "2021-11-25" @default.
- W4226523637 modified "2023-09-26" @default.
- W4226523637 title "ExPLoit: Extracting Private Labels in Split Learning" @default.
- W4226523637 doi "https://doi.org/10.48550/arxiv.2112.01299" @default.
- W4226523637 hasPublicationYear "2021" @default.
- W4226523637 type Work @default.
- W4226523637 citedByCount "0" @default.
- W4226523637 crossrefType "posted-content" @default.
- W4226523637 hasAuthorship W4226523637A5060352379 @default.
- W4226523637 hasAuthorship W4226523637A5082772077 @default.
- W4226523637 hasBestOaLocation W42265236371 @default.
- W4226523637 hasConcept C105795698 @default.
- W4226523637 hasConcept C119857082 @default.
- W4226523637 hasConcept C153180895 @default.
- W4226523637 hasConcept C154945302 @default.
- W4226523637 hasConcept C165064840 @default.
- W4226523637 hasConcept C165696696 @default.
- W4226523637 hasConcept C26517878 @default.
- W4226523637 hasConcept C2776135515 @default.
- W4226523637 hasConcept C33923547 @default.
- W4226523637 hasConcept C37736160 @default.
- W4226523637 hasConcept C38652104 @default.
- W4226523637 hasConcept C41008148 @default.
- W4226523637 hasConceptScore W4226523637C105795698 @default.
- W4226523637 hasConceptScore W4226523637C119857082 @default.
- W4226523637 hasConceptScore W4226523637C153180895 @default.
- W4226523637 hasConceptScore W4226523637C154945302 @default.
- W4226523637 hasConceptScore W4226523637C165064840 @default.
- W4226523637 hasConceptScore W4226523637C165696696 @default.
- W4226523637 hasConceptScore W4226523637C26517878 @default.
- W4226523637 hasConceptScore W4226523637C2776135515 @default.
- W4226523637 hasConceptScore W4226523637C33923547 @default.
- W4226523637 hasConceptScore W4226523637C37736160 @default.
- W4226523637 hasConceptScore W4226523637C38652104 @default.
- W4226523637 hasConceptScore W4226523637C41008148 @default.
- W4226523637 hasLocation W42265236371 @default.
- W4226523637 hasOpenAccess W4226523637 @default.
- W4226523637 hasPrimaryLocation W42265236371 @default.
- W4226523637 hasRelatedWork W2102298087 @default.
- W4226523637 hasRelatedWork W2603775267 @default.
- W4226523637 hasRelatedWork W2898291644 @default.
- W4226523637 hasRelatedWork W3033833039 @default.
- W4226523637 hasRelatedWork W3104224589 @default.
- W4226523637 hasRelatedWork W3124408655 @default.
- W4226523637 hasRelatedWork W4225854131 @default.
- W4226523637 hasRelatedWork W4297785512 @default.
- W4226523637 hasRelatedWork W4297845793 @default.
- W4226523637 hasRelatedWork W4379255972 @default.
- W4226523637 isParatext "false" @default.
- W4226523637 isRetracted "false" @default.
- W4226523637 workType "article" @default.