Matches in SemOpenAlex for { <https://semopenalex.org/work/W4309163068> ?p ?o ?g. }
Showing items 1 to 76 of
76
with 100 items per page.
- W4309163068 endingPage "2090" @default.
- W4309163068 startingPage "2081" @default.
- W4309163068 abstract "Dataset auditing for machine learning (ML) models is a method to evaluate if a given dataset is used in training a model. In a Federated Learning setting where multiple institutions collaboratively train a model with their decentralized private datasets, dataset auditing can facilitate the enforcement of regulations, which provide rules for preserving privacy, but also allow users to revoke authorizations and remove their data from collaboratively trained models. This paper first proposes a set of requirements for a practical dataset auditing method, and then present a novel dataset auditing method called Ensembled Membership Auditing ( <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink> <tex-math notation=LaTeX>$mathsf{EMA}$ </tex-math></inline-formula> ). Its key idea is to leverage previously proposed Membership Inference Attack methods and to aggregate data-wise membership scores using statistic testing to audit a dataset for a ML model. We have experimentally evaluated the proposed approach with benchmark datasets, as well as 4 X-ray datasets (CBIS-DDSM, COVIDx, Child-XRay, and CXR-NIH) and 3 dermatology datasets (DERM7pt, HAM10000, and PAD-UFES-20). Our results show that <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink> <tex-math notation=LaTeX>$mathsf{EMA}$ </tex-math></inline-formula> meet the requirements substantially better than the previous state-of-the-art method. Our code is at: <uri xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink>https://github.com/Hazelsuko07/EMA</uri> ." @default.
- W4309163068 created "2022-11-24" @default.
- W4309163068 creator A5051363890 @default.
- W4309163068 creator A5054972944 @default.
- W4309163068 creator A5082426637 @default.
- W4309163068 creator A5088533080 @default.
- W4309163068 date "2023-07-01" @default.
- W4309163068 modified "2023-10-17" @default.
- W4309163068 title "A Dataset Auditing Method for Collaboratively Trained Machine Learning Models" @default.
- W4309163068 cites W1488996941 @default.
- W4309163068 cites W2194775991 @default.
- W4309163068 cites W2535690855 @default.
- W4309163068 cites W2566079294 @default.
- W4309163068 cites W2757528734 @default.
- W4309163068 cites W2795435272 @default.
- W4309163068 cites W2797527544 @default.
- W4309163068 cites W2884943453 @default.
- W4309163068 cites W3029511006 @default.
- W4309163068 cites W3071470454 @default.
- W4309163068 cites W3081500256 @default.
- W4309163068 cites W3092391615 @default.
- W4309163068 cites W3101156210 @default.
- W4309163068 cites W3105081694 @default.
- W4309163068 cites W3154155772 @default.
- W4309163068 cites W3181414820 @default.
- W4309163068 cites W3203199349 @default.
- W4309163068 cites W3214586949 @default.
- W4309163068 doi "https://doi.org/10.1109/tmi.2022.3220706" @default.
- W4309163068 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/36378795" @default.
- W4309163068 hasPublicationYear "2023" @default.
- W4309163068 type Work @default.
- W4309163068 citedByCount "0" @default.
- W4309163068 crossrefType "journal-article" @default.
- W4309163068 hasAuthorship W4309163068A5051363890 @default.
- W4309163068 hasAuthorship W4309163068A5054972944 @default.
- W4309163068 hasAuthorship W4309163068A5082426637 @default.
- W4309163068 hasAuthorship W4309163068A5088533080 @default.
- W4309163068 hasConcept C119857082 @default.
- W4309163068 hasConcept C124101348 @default.
- W4309163068 hasConcept C153083717 @default.
- W4309163068 hasConcept C154945302 @default.
- W4309163068 hasConcept C162324750 @default.
- W4309163068 hasConcept C187736073 @default.
- W4309163068 hasConcept C199521495 @default.
- W4309163068 hasConcept C41008148 @default.
- W4309163068 hasConceptScore W4309163068C119857082 @default.
- W4309163068 hasConceptScore W4309163068C124101348 @default.
- W4309163068 hasConceptScore W4309163068C153083717 @default.
- W4309163068 hasConceptScore W4309163068C154945302 @default.
- W4309163068 hasConceptScore W4309163068C162324750 @default.
- W4309163068 hasConceptScore W4309163068C187736073 @default.
- W4309163068 hasConceptScore W4309163068C199521495 @default.
- W4309163068 hasConceptScore W4309163068C41008148 @default.
- W4309163068 hasFunder F4320315157 @default.
- W4309163068 hasFunder F4320334593 @default.
- W4309163068 hasIssue "7" @default.
- W4309163068 hasLocation W43091630681 @default.
- W4309163068 hasLocation W43091630682 @default.
- W4309163068 hasOpenAccess W4309163068 @default.
- W4309163068 hasPrimaryLocation W43091630681 @default.
- W4309163068 hasRelatedWork W2961085424 @default.
- W4309163068 hasRelatedWork W3046775127 @default.
- W4309163068 hasRelatedWork W3124999599 @default.
- W4309163068 hasRelatedWork W3170094116 @default.
- W4309163068 hasRelatedWork W4205958290 @default.
- W4309163068 hasRelatedWork W4285260836 @default.
- W4309163068 hasRelatedWork W4286629047 @default.
- W4309163068 hasRelatedWork W4306321456 @default.
- W4309163068 hasRelatedWork W4306674287 @default.
- W4309163068 hasRelatedWork W4224009465 @default.
- W4309163068 hasVolume "42" @default.
- W4309163068 isParatext "false" @default.
- W4309163068 isRetracted "false" @default.
- W4309163068 workType "article" @default.