Matches in SemOpenAlex for { <https://semopenalex.org/work/W3033258370> ?p ?o ?g. }
- W3033258370 abstract "Machine learning models are vulnerable to adversarial perturbations, that when added to an input, can cause high confidence misclassifications. The adversarial learning research community has made remarkable progress in the understanding of the root causes of adversarial perturbations. However, most problems that one may consider important to solve for the deployment of machine learning in safety critical tasks involve high dimensional complex manifolds that are difficult to characterize and study. It is common to develop adversarially robust learning theory on simple problems, in the hope that insights will transfer to `real world datasets'. In this work, we discuss a setting where this approach fails. In particular, we show with a linear classifier, it is always possible to solve a binary classification problem on Gaussian data under arbitrary levels of adversarial corruption during training, and that this property is not observed with non-linear classifiers on the CIFAR-10 dataset." @default.
- W3033258370 created "2020-06-12" @default.
- W3033258370 creator A5020715720 @default.
- W3033258370 date "2020-06-06" @default.
- W3033258370 modified "2023-09-27" @default.
- W3033258370 title "Unique properties of adversarially trained linear classifiers on Gaussian data." @default.
- W3033258370 cites W2194775991 @default.
- W3033258370 cites W2773446523 @default.
- W3033258370 cites W2799032899 @default.
- W3033258370 cites W2803732607 @default.
- W3033258370 cites W2807108439 @default.
- W3033258370 cites W2892179671 @default.
- W3033258370 cites W2897297967 @default.
- W3033258370 cites W2897355816 @default.
- W3033258370 cites W2898193427 @default.
- W3033258370 cites W2900604734 @default.
- W3033258370 cites W2906989593 @default.
- W3033258370 cites W2907518016 @default.
- W3033258370 cites W2913266441 @default.
- W3033258370 cites W2915022044 @default.
- W3033258370 cites W2950787360 @default.
- W3033258370 cites W2952049331 @default.
- W3033258370 cites W2952660984 @default.
- W3033258370 cites W2952847324 @default.
- W3033258370 cites W2962972504 @default.
- W3033258370 cites W2964153729 @default.
- W3033258370 cites W2964253222 @default.
- W3033258370 cites W2970209211 @default.
- W3033258370 cites W2970273067 @default.
- W3033258370 cites W2970680991 @default.
- W3033258370 cites W2970780203 @default.
- W3033258370 cites W2970864518 @default.
- W3033258370 cites W2971316968 @default.
- W3033258370 cites W2994164853 @default.
- W3033258370 cites W3006307757 @default.
- W3033258370 cites W3006935033 @default.
- W3033258370 cites W3007863415 @default.
- W3033258370 cites W3118608800 @default.
- W3033258370 hasPublicationYear "2020" @default.
- W3033258370 type Work @default.
- W3033258370 sameAs 3033258370 @default.
- W3033258370 citedByCount "0" @default.
- W3033258370 crossrefType "posted-content" @default.
- W3033258370 hasAuthorship W3033258370A5020715720 @default.
- W3033258370 hasConcept C111472728 @default.
- W3033258370 hasConcept C119857082 @default.
- W3033258370 hasConcept C121332964 @default.
- W3033258370 hasConcept C12267149 @default.
- W3033258370 hasConcept C138885662 @default.
- W3033258370 hasConcept C139532973 @default.
- W3033258370 hasConcept C154945302 @default.
- W3033258370 hasConcept C163716315 @default.
- W3033258370 hasConcept C189950617 @default.
- W3033258370 hasConcept C33923547 @default.
- W3033258370 hasConcept C37736160 @default.
- W3033258370 hasConcept C41008148 @default.
- W3033258370 hasConcept C48372109 @default.
- W3033258370 hasConcept C62520636 @default.
- W3033258370 hasConcept C66905080 @default.
- W3033258370 hasConcept C94375191 @default.
- W3033258370 hasConcept C95623464 @default.
- W3033258370 hasConceptScore W3033258370C111472728 @default.
- W3033258370 hasConceptScore W3033258370C119857082 @default.
- W3033258370 hasConceptScore W3033258370C121332964 @default.
- W3033258370 hasConceptScore W3033258370C12267149 @default.
- W3033258370 hasConceptScore W3033258370C138885662 @default.
- W3033258370 hasConceptScore W3033258370C139532973 @default.
- W3033258370 hasConceptScore W3033258370C154945302 @default.
- W3033258370 hasConceptScore W3033258370C163716315 @default.
- W3033258370 hasConceptScore W3033258370C189950617 @default.
- W3033258370 hasConceptScore W3033258370C33923547 @default.
- W3033258370 hasConceptScore W3033258370C37736160 @default.
- W3033258370 hasConceptScore W3033258370C41008148 @default.
- W3033258370 hasConceptScore W3033258370C48372109 @default.
- W3033258370 hasConceptScore W3033258370C62520636 @default.
- W3033258370 hasConceptScore W3033258370C66905080 @default.
- W3033258370 hasConceptScore W3033258370C94375191 @default.
- W3033258370 hasConceptScore W3033258370C95623464 @default.
- W3033258370 hasLocation W30332583701 @default.
- W3033258370 hasOpenAccess W3033258370 @default.
- W3033258370 hasPrimaryLocation W30332583701 @default.
- W3033258370 hasRelatedWork W2245869277 @default.
- W3033258370 hasRelatedWork W2781758978 @default.
- W3033258370 hasRelatedWork W2784285928 @default.
- W3033258370 hasRelatedWork W2884169041 @default.
- W3033258370 hasRelatedWork W2905582115 @default.
- W3033258370 hasRelatedWork W2963739340 @default.
- W3033258370 hasRelatedWork W2963753390 @default.
- W3033258370 hasRelatedWork W2963775761 @default.
- W3033258370 hasRelatedWork W2982324875 @default.
- W3033258370 hasRelatedWork W2995645057 @default.
- W3033258370 hasRelatedWork W2996687884 @default.
- W3033258370 hasRelatedWork W3003369812 @default.
- W3033258370 hasRelatedWork W3008051686 @default.
- W3033258370 hasRelatedWork W3018911559 @default.
- W3033258370 hasRelatedWork W3028525609 @default.
- W3033258370 hasRelatedWork W3036303027 @default.
- W3033258370 hasRelatedWork W3104032928 @default.
- W3033258370 hasRelatedWork W3122142687 @default.
- W3033258370 hasRelatedWork W3161366266 @default.