FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3199974954> ?p ?o ?g. }

• W3199974954 endingPage "7700" @default.
• W3199974954 startingPage "7691" @default.
• W3199974954 abstract "The right to erasure requires removal of a user's information from data held by organizations, with rigorous interpretations extending to downstream products such as learned models. Retraining from scratch with the particular user's data omitted fully removes its influence on the resulting model, but comes with a high computational cost. Machine unlearning mitigates the cost incurred by full retraining: instead, models are updated incrementally, possibly only requiring retraining when approximation errors accumulate. Rapid progress has been made towards privacy guarantees on the indistinguishability of unlearned and retrained models, but current formalisms do not place practical bounds on computation. In this paper we demonstrate how an attacker can exploit this oversight, highlighting a novel attack surface introduced by machine unlearning. We consider an attacker aiming to increase the computational cost of data removal. We derive and empirically investigate a poisoning attack on certified machine unlearning where strategically designed training data triggers complete retraining when removed." @default.
• W3199974954 created "2021-09-27" @default.
• W3199974954 creator A5047473680 @default.
• W3199974954 creator A5078824132 @default.
• W3199974954 creator A5080946416 @default.
• W3199974954 date "2022-06-28" @default.
• W3199974954 modified "2023-09-30" @default.
• W3199974954 title "Hard to Forget: Poisoning Attacks on Certified Machine Unlearning" @default.
• W3199974954 cites W134960717 @default.
• W3199974954 cites W1488996941 @default.
• W3199974954 cites W1557833142 @default.
• W3199974954 cites W1969811075 @default.
• W3199974954 cites W1978259121 @default.
• W3199974954 cites W1992129502 @default.
• W3199974954 cites W1994567799 @default.
• W3199974954 cites W2018711500 @default.
• W3199974954 cites W2051267297 @default.
• W3199974954 cites W2051434435 @default.
• W3199974954 cites W2095577883 @default.
• W3199974954 cites W2112507308 @default.
• W3199974954 cites W2112796928 @default.
• W3199974954 cites W2293844262 @default.
• W3199974954 cites W2535690855 @default.
• W3199974954 cites W2593892853 @default.
• W3199974954 cites W2597603852 @default.
• W3199974954 cites W2750384547 @default.
• W3199974954 cites W2774423163 @default.
• W3199974954 cites W2796004214 @default.
• W3199974954 cites W2798766386 @default.
• W3199974954 cites W2811973125 @default.
• W3199974954 cites W2888975495 @default.
• W3199974954 cites W2921723678 @default.
• W3199974954 cites W2946227741 @default.
• W3199974954 cites W2963207607 @default.
• W3199974954 cites W2963366334 @default.
• W3199974954 cites W2964043980 @default.
• W3199974954 cites W2964153729 @default.
• W3199974954 cites W2964253222 @default.
• W3199974954 cites W2971124187 @default.
• W3199974954 cites W2973217491 @default.
• W3199974954 cites W2983358147 @default.
• W3199974954 cites W2996800219 @default.
• W3199974954 cites W2999589176 @default.
• W3199974954 cites W3013335600 @default.
• W3199974954 cites W3014775455 @default.
• W3199974954 cites W3034412497 @default.
• W3199974954 cites W3035556513 @default.
• W3199974954 cites W3035644192 @default.
• W3199974954 cites W3040639636 @default.
• W3199974954 cites W3095509234 @default.
• W3199974954 cites W3119278161 @default.
• W3199974954 cites W3126379365 @default.
• W3199974954 cites W3127508141 @default.
• W3199974954 cites W3135378441 @default.
• W3199974954 cites W3154155772 @default.
• W3199974954 cites W3172075261 @default.
• W3199974954 cites W3176739818 @default.
• W3199974954 cites W3214586949 @default.
• W3199974954 doi "https://doi.org/10.1609/aaai.v36i7.20736" @default.
• W3199974954 hasPublicationYear "2022" @default.
• W3199974954 type Work @default.
• W3199974954 sameAs 3199974954 @default.
• W3199974954 citedByCount "6" @default.
• W3199974954 countsByYear W31999749542022 @default.
• W3199974954 countsByYear W31999749542023 @default.
• W3199974954 crossrefType "journal-article" @default.
• W3199974954 hasAuthorship W3199974954A5047473680 @default.
• W3199974954 hasAuthorship W3199974954A5078824132 @default.
• W3199974954 hasAuthorship W3199974954A5080946416 @default.
• W3199974954 hasBestOaLocation W31999749541 @default.
• W3199974954 hasConcept C107457646 @default.
• W3199974954 hasConcept C11413529 @default.
• W3199974954 hasConcept C119857082 @default.
• W3199974954 hasConcept C144133560 @default.
• W3199974954 hasConcept C154945302 @default.
• W3199974954 hasConcept C155202549 @default.
• W3199974954 hasConcept C165696696 @default.
• W3199974954 hasConcept C171018156 @default.
• W3199974954 hasConcept C17744445 @default.
• W3199974954 hasConcept C199360897 @default.
• W3199974954 hasConcept C199539241 @default.
• W3199974954 hasConcept C2524010 @default.
• W3199974954 hasConcept C2778712577 @default.
• W3199974954 hasConcept C2778790127 @default.
• W3199974954 hasConcept C33923547 @default.
• W3199974954 hasConcept C38652104 @default.
• W3199974954 hasConcept C41008148 @default.
• W3199974954 hasConcept C45374587 @default.
• W3199974954 hasConcept C46304622 @default.
• W3199974954 hasConceptScore W3199974954C107457646 @default.
• W3199974954 hasConceptScore W3199974954C11413529 @default.
• W3199974954 hasConceptScore W3199974954C119857082 @default.
• W3199974954 hasConceptScore W3199974954C144133560 @default.
• W3199974954 hasConceptScore W3199974954C154945302 @default.
• W3199974954 hasConceptScore W3199974954C155202549 @default.
• W3199974954 hasConceptScore W3199974954C165696696 @default.
• W3199974954 hasConceptScore W3199974954C171018156 @default.
• W3199974954 hasConceptScore W3199974954C17744445 @default.

• next