FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3171363095> ?p ?o ?g. }

• W3171363095 abstract "Deep Learning Accelerators are prone to faults which manifest in the form of errors in Neural Networks. Fault Tolerance in Neural Networks is crucial in real-time safety critical applications requiring computation for long durations. Neural Networks with high regularisation exhibit superior fault tolerance, however, at the cost of classification accuracy. In the view of difference in functionality, a Neural Network is modelled as two separate networks, i.e, the Feature Extractor with unsupervised learning objective and the Classifier with a supervised learning objective. Traditional approaches of training the entire network using a single supervised learning objective is insufficient to achieve the objectives of the individual components optimally. In this work, a novel multi-criteria objective function, combining unsupervised training of the Feature Extractor followed by supervised tuning with Classifier Network is proposed. The unsupervised training solves two games simultaneously in the presence of adversary neural networks with conflicting objectives to the Feature Extractor. The first game minimises the loss in reconstructing the input image for indistinguishability given the features from the Extractor, in the presence of a generative decoder. The second game solves a minimax constraint optimisation for distributional smoothening of feature space to match a prior distribution, in the presence of a Discriminator network. The resultant strongly regularised Feature Extractor is combined with the Classifier Network for supervised fine-tuning. The proposed Adversarial Fault Tolerant Neural Network Training is scalable to large networks and is independent of the architecture. The evaluation on benchmarking datasets: FashionMNIST and CIFAR10, indicates that the resultant networks have high accuracy with superior tolerance to stuck at 0 faults compared to widely used regularisers." @default.
• W3171363095 created "2021-06-22" @default.
• W3171363095 creator A5008638246 @default.
• W3171363095 creator A5037812896 @default.
• W3171363095 creator A5065664393 @default.
• W3171363095 date "2019-07-06" @default.
• W3171363095 modified "2023-09-27" @default.
• W3171363095 title "Towards Enhancing Fault Tolerance in Neural Networks" @default.
• W3171363095 cites W1987856988 @default.
• W3171363095 cites W2051513586 @default.
• W3171363095 cites W2098137815 @default.
• W3171363095 cites W2099471712 @default.
• W3171363095 cites W2111406701 @default.
• W3171363095 cites W2113502831 @default.
• W3171363095 cites W2123422798 @default.
• W3171363095 cites W2139072039 @default.
• W3171363095 cites W2139318413 @default.
• W3171363095 cites W2144784023 @default.
• W3171363095 cites W2145249131 @default.
• W3171363095 cites W2148605318 @default.
• W3171363095 cites W2149431208 @default.
• W3171363095 cites W2149593503 @default.
• W3171363095 cites W2152692509 @default.
• W3171363095 cites W2519070763 @default.
• W3171363095 cites W2589859332 @default.
• W3171363095 cites W2604319603 @default.
• W3171363095 cites W2606321545 @default.
• W3171363095 cites W2748528844 @default.
• W3171363095 cites W2749412917 @default.
• W3171363095 cites W2767260595 @default.
• W3171363095 cites W2894972989 @default.
• W3171363095 cites W2904958088 @default.
• W3171363095 cites W2945994301 @default.
• W3171363095 cites W2952825952 @default.
• W3171363095 cites W2963334011 @default.
• W3171363095 cites W2964866569 @default.
• W3171363095 cites W2989817299 @default.
• W3171363095 cites W3009818168 @default.
• W3171363095 hasPublicationYear "2019" @default.
• W3171363095 type Work @default.
• W3171363095 sameAs 3171363095 @default.
• W3171363095 citedByCount "0" @default.
• W3171363095 crossrefType "posted-content" @default.
• W3171363095 hasAuthorship W3171363095A5008638246 @default.
• W3171363095 hasAuthorship W3171363095A5037812896 @default.
• W3171363095 hasAuthorship W3171363095A5065664393 @default.
• W3171363095 hasConcept C108583219 @default.
• W3171363095 hasConcept C119857082 @default.
• W3171363095 hasConcept C153180895 @default.
• W3171363095 hasConcept C154945302 @default.
• W3171363095 hasConcept C2779803651 @default.
• W3171363095 hasConcept C41008148 @default.
• W3171363095 hasConcept C50644808 @default.
• W3171363095 hasConcept C76155785 @default.
• W3171363095 hasConcept C83665646 @default.
• W3171363095 hasConcept C94915269 @default.
• W3171363095 hasConcept C95623464 @default.
• W3171363095 hasConceptScore W3171363095C108583219 @default.
• W3171363095 hasConceptScore W3171363095C119857082 @default.
• W3171363095 hasConceptScore W3171363095C153180895 @default.
• W3171363095 hasConceptScore W3171363095C154945302 @default.
• W3171363095 hasConceptScore W3171363095C2779803651 @default.
• W3171363095 hasConceptScore W3171363095C41008148 @default.
• W3171363095 hasConceptScore W3171363095C50644808 @default.
• W3171363095 hasConceptScore W3171363095C76155785 @default.
• W3171363095 hasConceptScore W3171363095C83665646 @default.
• W3171363095 hasConceptScore W3171363095C94915269 @default.
• W3171363095 hasConceptScore W3171363095C95623464 @default.
• W3171363095 hasLocation W31713630951 @default.
• W3171363095 hasOpenAccess W3171363095 @default.
• W3171363095 hasPrimaryLocation W31713630951 @default.
• W3171363095 hasRelatedWork W1678611306 @default.
• W3171363095 hasRelatedWork W205793414 @default.
• W3171363095 hasRelatedWork W2077700203 @default.
• W3171363095 hasRelatedWork W2088572182 @default.
• W3171363095 hasRelatedWork W2103745410 @default.
• W3171363095 hasRelatedWork W2146555798 @default.
• W3171363095 hasRelatedWork W2540833927 @default.
• W3171363095 hasRelatedWork W2765284480 @default.
• W3171363095 hasRelatedWork W2774611252 @default.
• W3171363095 hasRelatedWork W2890821718 @default.
• W3171363095 hasRelatedWork W2894461463 @default.
• W3171363095 hasRelatedWork W2908935113 @default.
• W3171363095 hasRelatedWork W2954580816 @default.
• W3171363095 hasRelatedWork W3011102476 @default.
• W3171363095 hasRelatedWork W3016866710 @default.
• W3171363095 hasRelatedWork W3092600489 @default.
• W3171363095 hasRelatedWork W3184415262 @default.
• W3171363095 hasRelatedWork W3190315974 @default.
• W3171363095 hasRelatedWork W3205135396 @default.
• W3171363095 hasRelatedWork W1966635105 @default.
• W3171363095 isParatext "false" @default.
• W3171363095 isRetracted "false" @default.
• W3171363095 magId "3171363095" @default.
• W3171363095 workType "article" @default.