Matches in SemOpenAlex for { <https://semopenalex.org/work/W2894843516> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W2894843516 endingPage "359" @default.
- W2894843516 startingPage "346" @default.
- W2894843516 abstract "Abstract The model parameters of convolutional neural networks (CNNs) are determined by backpropagation (BP). In this work, we propose an interpretable feedforward (FF) design without any BP. The FF design adopts a data-centric approach. It derives network parameters of the current layer based on data statistics from the output of the previous layer in a one-pass manner. To construct convolutional layers, we develop a new signal transform, called the Saab (Subspace approximation with adjusted bias) transform. It is a variant of the principal component analysis with an added bias vector to annihilate activation’s nonlinearity. Multiple Saab transforms in cascade yield multiple convolutional layers. As to fully-connected layers, we construct them using a cascade of multi-stage linear least squared regressors. The classification and robustness performances of BP- and FF-designed CNNs applied to the MNIST and the CIFAR-10 datasets are compared. Finally, we comment on the relationship between BP and FF designs." @default.
- W2894843516 created "2018-10-12" @default.
- W2894843516 creator A5001082656 @default.
- W2894843516 creator A5008019088 @default.
- W2894843516 creator A5009039461 @default.
- W2894843516 creator A5010841520 @default.
- W2894843516 creator A5065995768 @default.
- W2894843516 date "2019-04-01" @default.
- W2894843516 modified "2023-10-06" @default.
- W2894843516 title "Interpretable convolutional neural networks via feedforward design" @default.
- W2894843516 cites W1787224781 @default.
- W2894843516 cites W1994906459 @default.
- W2894843516 cites W2072072671 @default.
- W2894843516 cites W2073681337 @default.
- W2894843516 cites W2103496339 @default.
- W2894843516 cites W2112796928 @default.
- W2894843516 cites W2137983211 @default.
- W2894843516 cites W2146395539 @default.
- W2894843516 cites W2149194912 @default.
- W2894843516 cites W2519796145 @default.
- W2894843516 cites W2608422103 @default.
- W2894843516 cites W2903823300 @default.
- W2894843516 cites W2962690605 @default.
- W2894843516 cites W2963417959 @default.
- W2894843516 doi "https://doi.org/10.1016/j.jvcir.2019.03.010" @default.
- W2894843516 hasPublicationYear "2019" @default.
- W2894843516 type Work @default.
- W2894843516 sameAs 2894843516 @default.
- W2894843516 citedByCount "108" @default.
- W2894843516 countsByYear W28948435162018 @default.
- W2894843516 countsByYear W28948435162019 @default.
- W2894843516 countsByYear W28948435162020 @default.
- W2894843516 countsByYear W28948435162021 @default.
- W2894843516 countsByYear W28948435162022 @default.
- W2894843516 countsByYear W28948435162023 @default.
- W2894843516 crossrefType "journal-article" @default.
- W2894843516 hasAuthorship W2894843516A5001082656 @default.
- W2894843516 hasAuthorship W2894843516A5008019088 @default.
- W2894843516 hasAuthorship W2894843516A5009039461 @default.
- W2894843516 hasAuthorship W2894843516A5010841520 @default.
- W2894843516 hasAuthorship W2894843516A5065995768 @default.
- W2894843516 hasBestOaLocation W28948435162 @default.
- W2894843516 hasConcept C127413603 @default.
- W2894843516 hasConcept C133731056 @default.
- W2894843516 hasConcept C153180895 @default.
- W2894843516 hasConcept C154945302 @default.
- W2894843516 hasConcept C28490314 @default.
- W2894843516 hasConcept C38858127 @default.
- W2894843516 hasConcept C41008148 @default.
- W2894843516 hasConcept C47702885 @default.
- W2894843516 hasConcept C50644808 @default.
- W2894843516 hasConcept C81363708 @default.
- W2894843516 hasConceptScore W2894843516C127413603 @default.
- W2894843516 hasConceptScore W2894843516C133731056 @default.
- W2894843516 hasConceptScore W2894843516C153180895 @default.
- W2894843516 hasConceptScore W2894843516C154945302 @default.
- W2894843516 hasConceptScore W2894843516C28490314 @default.
- W2894843516 hasConceptScore W2894843516C38858127 @default.
- W2894843516 hasConceptScore W2894843516C41008148 @default.
- W2894843516 hasConceptScore W2894843516C47702885 @default.
- W2894843516 hasConceptScore W2894843516C50644808 @default.
- W2894843516 hasConceptScore W2894843516C81363708 @default.
- W2894843516 hasLocation W28948435161 @default.
- W2894843516 hasLocation W28948435162 @default.
- W2894843516 hasOpenAccess W2894843516 @default.
- W2894843516 hasPrimaryLocation W28948435161 @default.
- W2894843516 hasRelatedWork W1604847762 @default.
- W2894843516 hasRelatedWork W2086999410 @default.
- W2894843516 hasRelatedWork W2175746458 @default.
- W2894843516 hasRelatedWork W2258992572 @default.
- W2894843516 hasRelatedWork W2338394561 @default.
- W2894843516 hasRelatedWork W2359410228 @default.
- W2894843516 hasRelatedWork W2732542196 @default.
- W2894843516 hasRelatedWork W2889587233 @default.
- W2894843516 hasRelatedWork W3081496756 @default.
- W2894843516 hasRelatedWork W3093612317 @default.
- W2894843516 hasVolume "60" @default.
- W2894843516 isParatext "false" @default.
- W2894843516 isRetracted "false" @default.
- W2894843516 magId "2894843516" @default.
- W2894843516 workType "article" @default.