Matches in SemOpenAlex for { <https://semopenalex.org/work/W3160386323> ?p ?o ?g. }
Showing items 1 to 99 of
99
with 100 items per page.
- W3160386323 abstract "In this work we unify a number of inference learning methods, that are proposed in the literature as alternative training algorithms to the ones based on regular error back-propagation. These inference learning methods were developed with very diverse motivations, mainly aiming to enhance the biological plausibility of deep neural networks and to improve the intrinsic parallelism of training methods. We show that these superficially very different methods can all be obtained by successively applying a particular reformulation of bilevel optimization programs. As a by-product it becomes also evident that all considered inference learning methods include back-propagation as a special case, and therefore at least approximate error back-propagation in typical settings. Finally, we propose Fenchel back-propagation, that replaces the propagation of infinitesimal corrections performed in standard back-propagation with finite targets as the learning signal. Fenchel back-propagation can therefore be seen as an instance of learning via explicit target propagation." @default.
- W3160386323 created "2021-05-24" @default.
- W3160386323 creator A5020965719 @default.
- W3160386323 date "2021-05-15" @default.
- W3160386323 modified "2023-09-27" @default.
- W3160386323 title "Bilevel Programming and Deep Learning: A Unifying View on Inference Learning Methods." @default.
- W3160386323 cites W1492899839 @default.
- W3160386323 cites W1498436455 @default.
- W3160386323 cites W1520121841 @default.
- W3160386323 cites W1597009674 @default.
- W3160386323 cites W1606458877 @default.
- W3160386323 cites W1881179843 @default.
- W3160386323 cites W1970163214 @default.
- W3160386323 cites W2007431958 @default.
- W3160386323 cites W2069900289 @default.
- W3160386323 cites W2102686659 @default.
- W3160386323 cites W2116341502 @default.
- W3160386323 cites W2134664762 @default.
- W3160386323 cites W2150355110 @default.
- W3160386323 cites W2156417353 @default.
- W3160386323 cites W2163112044 @default.
- W3160386323 cites W2166107799 @default.
- W3160386323 cites W2187061624 @default.
- W3160386323 cites W2278603371 @default.
- W3160386323 cites W2395484609 @default.
- W3160386323 cites W2527798464 @default.
- W3160386323 cites W2604086949 @default.
- W3160386323 cites W2799577210 @default.
- W3160386323 cites W2962985678 @default.
- W3160386323 cites W2963055445 @default.
- W3160386323 cites W2963233958 @default.
- W3160386323 cites W2963306862 @default.
- W3160386323 cites W2963830987 @default.
- W3160386323 cites W2963904364 @default.
- W3160386323 cites W2964121744 @default.
- W3160386323 cites W2972954809 @default.
- W3160386323 cites W3023420017 @default.
- W3160386323 cites W3033108376 @default.
- W3160386323 cites W3037120332 @default.
- W3160386323 cites W3037315641 @default.
- W3160386323 cites W3037884961 @default.
- W3160386323 cites W3100445868 @default.
- W3160386323 cites W3113907021 @default.
- W3160386323 cites W3171045091 @default.
- W3160386323 hasPublicationYear "2021" @default.
- W3160386323 type Work @default.
- W3160386323 sameAs 3160386323 @default.
- W3160386323 citedByCount "0" @default.
- W3160386323 crossrefType "posted-content" @default.
- W3160386323 hasAuthorship W3160386323A5020965719 @default.
- W3160386323 hasConcept C108583219 @default.
- W3160386323 hasConcept C11413529 @default.
- W3160386323 hasConcept C119857082 @default.
- W3160386323 hasConcept C137836250 @default.
- W3160386323 hasConcept C154945302 @default.
- W3160386323 hasConcept C155032097 @default.
- W3160386323 hasConcept C2776214188 @default.
- W3160386323 hasConcept C2777472644 @default.
- W3160386323 hasConcept C3309286 @default.
- W3160386323 hasConcept C41008148 @default.
- W3160386323 hasConcept C50644808 @default.
- W3160386323 hasConceptScore W3160386323C108583219 @default.
- W3160386323 hasConceptScore W3160386323C11413529 @default.
- W3160386323 hasConceptScore W3160386323C119857082 @default.
- W3160386323 hasConceptScore W3160386323C137836250 @default.
- W3160386323 hasConceptScore W3160386323C154945302 @default.
- W3160386323 hasConceptScore W3160386323C155032097 @default.
- W3160386323 hasConceptScore W3160386323C2776214188 @default.
- W3160386323 hasConceptScore W3160386323C2777472644 @default.
- W3160386323 hasConceptScore W3160386323C3309286 @default.
- W3160386323 hasConceptScore W3160386323C41008148 @default.
- W3160386323 hasConceptScore W3160386323C50644808 @default.
- W3160386323 hasLocation W31603863231 @default.
- W3160386323 hasOpenAccess W3160386323 @default.
- W3160386323 hasPrimaryLocation W31603863231 @default.
- W3160386323 hasRelatedWork W2043700026 @default.
- W3160386323 hasRelatedWork W2182299232 @default.
- W3160386323 hasRelatedWork W2345689272 @default.
- W3160386323 hasRelatedWork W2607462057 @default.
- W3160386323 hasRelatedWork W2626454479 @default.
- W3160386323 hasRelatedWork W2683470288 @default.
- W3160386323 hasRelatedWork W2727634334 @default.
- W3160386323 hasRelatedWork W2799053100 @default.
- W3160386323 hasRelatedWork W2803895400 @default.
- W3160386323 hasRelatedWork W2963266548 @default.
- W3160386323 hasRelatedWork W2963863953 @default.
- W3160386323 hasRelatedWork W2977824741 @default.
- W3160386323 hasRelatedWork W2991051561 @default.
- W3160386323 hasRelatedWork W3011241234 @default.
- W3160386323 hasRelatedWork W3017302611 @default.
- W3160386323 hasRelatedWork W3039808591 @default.
- W3160386323 hasRelatedWork W3124367822 @default.
- W3160386323 hasRelatedWork W3174659074 @default.
- W3160386323 hasRelatedWork W3198407848 @default.
- W3160386323 hasRelatedWork W2186947120 @default.
- W3160386323 isParatext "false" @default.
- W3160386323 isRetracted "false" @default.
- W3160386323 magId "3160386323" @default.
- W3160386323 workType "article" @default.