Matches in SemOpenAlex for { <https://semopenalex.org/work/W2913608505> ?p ?o ?g. }
Showing items 1 to 78 of
78
with 100 items per page.
- W2913608505 abstract "As online systems based on machine learning are offered to public or paid subscribers via application programming interfaces (APIs), they become vulnerable to frequent exploits and attacks. This paper studies adversarial machine learning in the practical case when there are rate limitations on API calls. The adversary launches an exploratory (inference) attack by querying the API of an online machine learning system (in particular, a classifier) with input data samples, collecting returned labels to build up the training data, and training an adversarial classifier that is functionally equivalent and statistically close to the target classifier. The exploratory attack with limited training data is shown to fail to reliably infer the target classifier of a real text classifier API that is available online to the public. In return, a generative adversarial network (GAN) based on deep learning is built to generate synthetic training data from a limited number of real training data samples, thereby extending the training data and improving the performance of the inferred classifier. The exploratory attack provides the basis to launch the causative attack (that aims to poison the training process) and evasion attack (that aims to fool the classifier into making wrong decisions) by selecting training and test data samples, respectively, based on the confidence scores obtained from the inferred classifier. These stealth attacks with small footprint (using a small number of API calls) make adversarial machine learning practical under the realistic case with limited training data available to the adversary." @default.
- W2913608505 created "2019-02-21" @default.
- W2913608505 creator A5054863127 @default.
- W2913608505 creator A5075186616 @default.
- W2913608505 creator A5075960743 @default.
- W2913608505 creator A5089570143 @default.
- W2913608505 date "2018-12-01" @default.
- W2913608505 modified "2023-10-16" @default.
- W2913608505 title "Generative Adversarial Networks for Black-Box API Attacks with Limited Training Data" @default.
- W2913608505 cites W2151298633 @default.
- W2913608505 cites W2180612164 @default.
- W2913608505 cites W2512472178 @default.
- W2913608505 cites W2610603914 @default.
- W2913608505 cites W2623427976 @default.
- W2913608505 cites W2773459750 @default.
- W2913608505 cites W2963079272 @default.
- W2913608505 cites W2963162645 @default.
- W2913608505 doi "https://doi.org/10.1109/isspit.2018.8642683" @default.
- W2913608505 hasPublicationYear "2018" @default.
- W2913608505 type Work @default.
- W2913608505 sameAs 2913608505 @default.
- W2913608505 citedByCount "25" @default.
- W2913608505 countsByYear W29136085052018 @default.
- W2913608505 countsByYear W29136085052019 @default.
- W2913608505 countsByYear W29136085052020 @default.
- W2913608505 countsByYear W29136085052021 @default.
- W2913608505 countsByYear W29136085052022 @default.
- W2913608505 countsByYear W29136085052023 @default.
- W2913608505 crossrefType "proceedings-article" @default.
- W2913608505 hasAuthorship W2913608505A5054863127 @default.
- W2913608505 hasAuthorship W2913608505A5075186616 @default.
- W2913608505 hasAuthorship W2913608505A5075960743 @default.
- W2913608505 hasAuthorship W2913608505A5089570143 @default.
- W2913608505 hasBestOaLocation W29136085052 @default.
- W2913608505 hasConcept C108583219 @default.
- W2913608505 hasConcept C119857082 @default.
- W2913608505 hasConcept C154945302 @default.
- W2913608505 hasConcept C165696696 @default.
- W2913608505 hasConcept C2776214188 @default.
- W2913608505 hasConcept C2778403875 @default.
- W2913608505 hasConcept C2988773926 @default.
- W2913608505 hasConcept C37736160 @default.
- W2913608505 hasConcept C38652104 @default.
- W2913608505 hasConcept C39890363 @default.
- W2913608505 hasConcept C41008148 @default.
- W2913608505 hasConcept C41065033 @default.
- W2913608505 hasConcept C95623464 @default.
- W2913608505 hasConceptScore W2913608505C108583219 @default.
- W2913608505 hasConceptScore W2913608505C119857082 @default.
- W2913608505 hasConceptScore W2913608505C154945302 @default.
- W2913608505 hasConceptScore W2913608505C165696696 @default.
- W2913608505 hasConceptScore W2913608505C2776214188 @default.
- W2913608505 hasConceptScore W2913608505C2778403875 @default.
- W2913608505 hasConceptScore W2913608505C2988773926 @default.
- W2913608505 hasConceptScore W2913608505C37736160 @default.
- W2913608505 hasConceptScore W2913608505C38652104 @default.
- W2913608505 hasConceptScore W2913608505C39890363 @default.
- W2913608505 hasConceptScore W2913608505C41008148 @default.
- W2913608505 hasConceptScore W2913608505C41065033 @default.
- W2913608505 hasConceptScore W2913608505C95623464 @default.
- W2913608505 hasLocation W29136085051 @default.
- W2913608505 hasLocation W29136085052 @default.
- W2913608505 hasOpenAccess W2913608505 @default.
- W2913608505 hasPrimaryLocation W29136085051 @default.
- W2913608505 hasRelatedWork W2787496614 @default.
- W2913608505 hasRelatedWork W2891797126 @default.
- W2913608505 hasRelatedWork W2964197269 @default.
- W2913608505 hasRelatedWork W2974723675 @default.
- W2913608505 hasRelatedWork W3013617128 @default.
- W2913608505 hasRelatedWork W3124408655 @default.
- W2913608505 hasRelatedWork W3156291593 @default.
- W2913608505 hasRelatedWork W4297785512 @default.
- W2913608505 hasRelatedWork W4310580317 @default.
- W2913608505 hasRelatedWork W4380925710 @default.
- W2913608505 isParatext "false" @default.
- W2913608505 isRetracted "false" @default.
- W2913608505 magId "2913608505" @default.
- W2913608505 workType "article" @default.