Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385955293> ?p ?o ?g. }
Showing items 1 to 78 of
78
with 100 items per page.
- W4385955293 endingPage "9341" @default.
- W4385955293 startingPage "9341" @default.
- W4385955293 abstract "Recent legislations, such as the European General Data Protection Regulation (GDPR), require user data holders to guarantee the individual’s right to be forgotten. This means that user data holders must completely delete user data upon request. However, in the field of machine learning, it is not possible to simply remove these data from the back-end database wherein the training dataset is stored, because the machine learning model still retains this data information. Retraining the model using a dataset with these data removed can overcome this problem; however, this can lead to expensive computational overheads. In order to remedy this shortcoming, we propose two effective methods to help model owners or data holders remove private data from a trained model. The first method uses an elastic weight consolidation (EWC) constraint term and a modified loss function to neutralize the data to be removed. The second method approximates the posterior distribution of the model as a Gaussian distribution, and the model after unlearning is computed by decreasingly matching the moment (DMM) of the posterior distribution of the neural network trained on all data and the data to be removed. Finally, we conducted experiments on three standard datasets using backdoor attacks as the evaluation metric. The results show that both methods are effective in removing backdoor triggers in deep learning models. Specifically, EWC can reduce the success rate of backdoor attacks to 0. IMM can ensure that the model prediction accuracy is higher than 80% and keep the success rate of backdoor attacks below 10%." @default.
- W4385955293 created "2023-08-18" @default.
- W4385955293 creator A5034852123 @default.
- W4385955293 creator A5044705390 @default.
- W4385955293 creator A5058533368 @default.
- W4385955293 creator A5059682045 @default.
- W4385955293 creator A5070452233 @default.
- W4385955293 date "2023-08-17" @default.
- W4385955293 modified "2023-10-17" @default.
- W4385955293 title "Machine Unlearning by Reversing the Continual Learning" @default.
- W4385955293 cites W1488996941 @default.
- W4385955293 cites W1682403713 @default.
- W4385955293 cites W1981184437 @default.
- W4385955293 cites W2194775991 @default.
- W4385955293 cites W2535690855 @default.
- W4385955293 cites W2560647685 @default.
- W4385955293 cites W3004443750 @default.
- W4385955293 cites W3017285899 @default.
- W4385955293 cites W3035644192 @default.
- W4385955293 cites W3135028703 @default.
- W4385955293 cites W3154155772 @default.
- W4385955293 cites W3202838631 @default.
- W4385955293 cites W3209576002 @default.
- W4385955293 cites W3210628769 @default.
- W4385955293 cites W3214586949 @default.
- W4385955293 cites W4250732133 @default.
- W4385955293 cites W4292260100 @default.
- W4385955293 cites W4292793961 @default.
- W4385955293 cites W4321615196 @default.
- W4385955293 doi "https://doi.org/10.3390/app13169341" @default.
- W4385955293 hasPublicationYear "2023" @default.
- W4385955293 type Work @default.
- W4385955293 citedByCount "0" @default.
- W4385955293 crossrefType "journal-article" @default.
- W4385955293 hasAuthorship W4385955293A5034852123 @default.
- W4385955293 hasAuthorship W4385955293A5044705390 @default.
- W4385955293 hasAuthorship W4385955293A5058533368 @default.
- W4385955293 hasAuthorship W4385955293A5059682045 @default.
- W4385955293 hasAuthorship W4385955293A5070452233 @default.
- W4385955293 hasBestOaLocation W43859552931 @default.
- W4385955293 hasConcept C119857082 @default.
- W4385955293 hasConcept C124101348 @default.
- W4385955293 hasConcept C144133560 @default.
- W4385955293 hasConcept C154945302 @default.
- W4385955293 hasConcept C155202549 @default.
- W4385955293 hasConcept C2778712577 @default.
- W4385955293 hasConcept C2781045450 @default.
- W4385955293 hasConcept C38652104 @default.
- W4385955293 hasConcept C41008148 @default.
- W4385955293 hasConceptScore W4385955293C119857082 @default.
- W4385955293 hasConceptScore W4385955293C124101348 @default.
- W4385955293 hasConceptScore W4385955293C144133560 @default.
- W4385955293 hasConceptScore W4385955293C154945302 @default.
- W4385955293 hasConceptScore W4385955293C155202549 @default.
- W4385955293 hasConceptScore W4385955293C2778712577 @default.
- W4385955293 hasConceptScore W4385955293C2781045450 @default.
- W4385955293 hasConceptScore W4385955293C38652104 @default.
- W4385955293 hasConceptScore W4385955293C41008148 @default.
- W4385955293 hasIssue "16" @default.
- W4385955293 hasLocation W43859552931 @default.
- W4385955293 hasOpenAccess W4385955293 @default.
- W4385955293 hasPrimaryLocation W43859552931 @default.
- W4385955293 hasRelatedWork W2961085424 @default.
- W4385955293 hasRelatedWork W3046775127 @default.
- W4385955293 hasRelatedWork W3107602296 @default.
- W4385955293 hasRelatedWork W3170094116 @default.
- W4385955293 hasRelatedWork W3209574120 @default.
- W4385955293 hasRelatedWork W4210805261 @default.
- W4385955293 hasRelatedWork W4306674287 @default.
- W4385955293 hasRelatedWork W4312192474 @default.
- W4385955293 hasRelatedWork W4386462264 @default.
- W4385955293 hasRelatedWork W4387297750 @default.
- W4385955293 hasVolume "13" @default.
- W4385955293 isParatext "false" @default.
- W4385955293 isRetracted "false" @default.
- W4385955293 workType "article" @default.