Matches in SemOpenAlex for { <https://semopenalex.org/work/W2799132109> ?p ?o ?g. }
Showing items 1 to 96 of
96
with 100 items per page.
- W2799132109 abstract "Convolutional and Recurrent, deep neural networks have been successful in machine learning systems for computer vision, reinforcement learning, and other allied fields. However, the robustness of such neural networks is seldom apprised, especially after high classification accuracy has been attained. In this paper, we evaluate the robustness of three recurrent neural networks to tiny perturbations, on three widely used datasets, to argue that high accuracy does not always mean a stable and a robust (to bounded perturbations, adversarial attacks, etc.) system. Especially, normalizing the spectrum of the discrete recurrent network to bound the spectrum (using power method, Rayleigh quotient, etc.) on a unit disk produces stable, albeit highly non-robust neural networks. Furthermore, using the $epsilon$-pseudo-spectrum, we show that training of recurrent networks, say using gradient-based methods, often result in non-normal matrices that may or may not be diagonalizable. Therefore, the open problem lies in constructing methods that optimize not only for accuracy but also for the stability and the robustness of the underlying neural network, a criterion that is distinct from the other." @default.
- W2799132109 created "2018-05-07" @default.
- W2799132109 creator A5055824933 @default.
- W2799132109 creator A5077250049 @default.
- W2799132109 date "2018-04-30" @default.
- W2799132109 modified "2023-09-27" @default.
- W2799132109 title "How Robust are Deep Neural Networks" @default.
- W2799132109 cites W1591706642 @default.
- W2799132109 cites W1968315580 @default.
- W2799132109 cites W1998151455 @default.
- W2799132109 cites W2014076462 @default.
- W2799132109 cites W2064675550 @default.
- W2799132109 cites W2087301801 @default.
- W2799132109 cites W2104681180 @default.
- W2799132109 cites W2105398397 @default.
- W2799132109 cites W2107280169 @default.
- W2799132109 cites W2113459411 @default.
- W2799132109 cites W2148242024 @default.
- W2799132109 cites W2293942407 @default.
- W2799132109 cites W2587753047 @default.
- W2799132109 cites W2593744649 @default.
- W2799132109 cites W2753698329 @default.
- W2799132109 cites W2767798701 @default.
- W2799132109 cites W2798065830 @default.
- W2799132109 cites W2950527759 @default.
- W2799132109 cites W2951605425 @default.
- W2799132109 cites W2963042606 @default.
- W2799132109 cites W2963046515 @default.
- W2799132109 hasPublicationYear "2018" @default.
- W2799132109 type Work @default.
- W2799132109 sameAs 2799132109 @default.
- W2799132109 citedByCount "3" @default.
- W2799132109 countsByYear W27991321092019 @default.
- W2799132109 countsByYear W27991321092020 @default.
- W2799132109 crossrefType "posted-content" @default.
- W2799132109 hasAuthorship W2799132109A5055824933 @default.
- W2799132109 hasAuthorship W2799132109A5077250049 @default.
- W2799132109 hasConcept C104317684 @default.
- W2799132109 hasConcept C108583219 @default.
- W2799132109 hasConcept C11413529 @default.
- W2799132109 hasConcept C119857082 @default.
- W2799132109 hasConcept C134306372 @default.
- W2799132109 hasConcept C147168706 @default.
- W2799132109 hasConcept C154945302 @default.
- W2799132109 hasConcept C185592680 @default.
- W2799132109 hasConcept C2984842247 @default.
- W2799132109 hasConcept C33923547 @default.
- W2799132109 hasConcept C34388435 @default.
- W2799132109 hasConcept C41008148 @default.
- W2799132109 hasConcept C50644808 @default.
- W2799132109 hasConcept C55493867 @default.
- W2799132109 hasConcept C63479239 @default.
- W2799132109 hasConcept C81363708 @default.
- W2799132109 hasConceptScore W2799132109C104317684 @default.
- W2799132109 hasConceptScore W2799132109C108583219 @default.
- W2799132109 hasConceptScore W2799132109C11413529 @default.
- W2799132109 hasConceptScore W2799132109C119857082 @default.
- W2799132109 hasConceptScore W2799132109C134306372 @default.
- W2799132109 hasConceptScore W2799132109C147168706 @default.
- W2799132109 hasConceptScore W2799132109C154945302 @default.
- W2799132109 hasConceptScore W2799132109C185592680 @default.
- W2799132109 hasConceptScore W2799132109C2984842247 @default.
- W2799132109 hasConceptScore W2799132109C33923547 @default.
- W2799132109 hasConceptScore W2799132109C34388435 @default.
- W2799132109 hasConceptScore W2799132109C41008148 @default.
- W2799132109 hasConceptScore W2799132109C50644808 @default.
- W2799132109 hasConceptScore W2799132109C55493867 @default.
- W2799132109 hasConceptScore W2799132109C63479239 @default.
- W2799132109 hasConceptScore W2799132109C81363708 @default.
- W2799132109 hasLocation W27991321091 @default.
- W2799132109 hasOpenAccess W2799132109 @default.
- W2799132109 hasPrimaryLocation W27991321091 @default.
- W2799132109 hasRelatedWork W2095705004 @default.
- W2799132109 hasRelatedWork W2113517874 @default.
- W2799132109 hasRelatedWork W2163605009 @default.
- W2799132109 hasRelatedWork W2194775991 @default.
- W2799132109 hasRelatedWork W2342045095 @default.
- W2799132109 hasRelatedWork W2463256602 @default.
- W2799132109 hasRelatedWork W2609644570 @default.
- W2799132109 hasRelatedWork W2742801844 @default.
- W2799132109 hasRelatedWork W2746420172 @default.
- W2799132109 hasRelatedWork W2761490953 @default.
- W2799132109 hasRelatedWork W2785333883 @default.
- W2799132109 hasRelatedWork W2806294667 @default.
- W2799132109 hasRelatedWork W2919115771 @default.
- W2799132109 hasRelatedWork W2964153729 @default.
- W2799132109 hasRelatedWork W2987612595 @default.
- W2799132109 hasRelatedWork W3032366304 @default.
- W2799132109 hasRelatedWork W3041782238 @default.
- W2799132109 hasRelatedWork W3098011980 @default.
- W2799132109 hasRelatedWork W3193683518 @default.
- W2799132109 hasRelatedWork W3003520923 @default.
- W2799132109 isParatext "false" @default.
- W2799132109 isRetracted "false" @default.
- W2799132109 magId "2799132109" @default.
- W2799132109 workType "article" @default.