Matches in SemOpenAlex for { <https://semopenalex.org/work/W3085710185> ?p ?o ?g. }
Showing items 1 to 90 of
90
with 100 items per page.
- W3085710185 endingPage "38" @default.
- W3085710185 startingPage "1" @default.
- W3085710185 abstract "In this study, we prove that an intrinsic low dimensionality of covariates is the main factor that determines the performance of deep neural networks (DNNs). DNNs generally provide outstanding empirical performance. Hence, numerous studies have actively investigated the theoretical properties of DNNs to understand their underlying mechanisms. In particular, the behavior of DNNs in terms of high-dimensional data is one of the most critical questions. However, this issue has not been sufficiently investigated from the aspect of covariates, although high-dimensional data have practically low intrinsic dimensionality. In this study, we derive bounds for an approximation error and a generalization error regarding DNNs with intrinsically low dimensional covariates. We apply the notion of the Minkowski dimension and develop a novel proof technique. Consequently, we show that convergence rates of the errors by DNNs do not depend on the nominal high dimensionality of data, but on its lower intrinsic dimension. We further prove that the rate is optimal in the minimax sense. We identify an advantage of DNNs by showing that DNNs can handle a broader class of intrinsic low dimensional data than other adaptive estimators. Finally, we conduct a numerical simulation to validate the theoretical results." @default.
- W3085710185 created "2020-09-21" @default.
- W3085710185 creator A5066308746 @default.
- W3085710185 creator A5088508553 @default.
- W3085710185 date "2020-01-01" @default.
- W3085710185 modified "2023-09-26" @default.
- W3085710185 title "Adaptive Approximation and Generalization of Deep Neural Network with Intrinsic Dimensionality" @default.
- W3085710185 hasPublicationYear "2020" @default.
- W3085710185 type Work @default.
- W3085710185 sameAs 3085710185 @default.
- W3085710185 citedByCount "14" @default.
- W3085710185 countsByYear W30857101852019 @default.
- W3085710185 countsByYear W30857101852020 @default.
- W3085710185 countsByYear W30857101852021 @default.
- W3085710185 crossrefType "journal-article" @default.
- W3085710185 hasAuthorship W3085710185A5066308746 @default.
- W3085710185 hasAuthorship W3085710185A5088508553 @default.
- W3085710185 hasConcept C105795698 @default.
- W3085710185 hasConcept C111030470 @default.
- W3085710185 hasConcept C119043178 @default.
- W3085710185 hasConcept C119857082 @default.
- W3085710185 hasConcept C126255220 @default.
- W3085710185 hasConcept C134306372 @default.
- W3085710185 hasConcept C149728462 @default.
- W3085710185 hasConcept C154945302 @default.
- W3085710185 hasConcept C162324750 @default.
- W3085710185 hasConcept C177148314 @default.
- W3085710185 hasConcept C185429906 @default.
- W3085710185 hasConcept C202444582 @default.
- W3085710185 hasConcept C2777303404 @default.
- W3085710185 hasConcept C28826006 @default.
- W3085710185 hasConcept C2984842247 @default.
- W3085710185 hasConcept C30732413 @default.
- W3085710185 hasConcept C33676613 @default.
- W3085710185 hasConcept C33923547 @default.
- W3085710185 hasConcept C41008148 @default.
- W3085710185 hasConcept C50522688 @default.
- W3085710185 hasConcept C50644808 @default.
- W3085710185 hasConceptScore W3085710185C105795698 @default.
- W3085710185 hasConceptScore W3085710185C111030470 @default.
- W3085710185 hasConceptScore W3085710185C119043178 @default.
- W3085710185 hasConceptScore W3085710185C119857082 @default.
- W3085710185 hasConceptScore W3085710185C126255220 @default.
- W3085710185 hasConceptScore W3085710185C134306372 @default.
- W3085710185 hasConceptScore W3085710185C149728462 @default.
- W3085710185 hasConceptScore W3085710185C154945302 @default.
- W3085710185 hasConceptScore W3085710185C162324750 @default.
- W3085710185 hasConceptScore W3085710185C177148314 @default.
- W3085710185 hasConceptScore W3085710185C185429906 @default.
- W3085710185 hasConceptScore W3085710185C202444582 @default.
- W3085710185 hasConceptScore W3085710185C2777303404 @default.
- W3085710185 hasConceptScore W3085710185C28826006 @default.
- W3085710185 hasConceptScore W3085710185C2984842247 @default.
- W3085710185 hasConceptScore W3085710185C30732413 @default.
- W3085710185 hasConceptScore W3085710185C33676613 @default.
- W3085710185 hasConceptScore W3085710185C33923547 @default.
- W3085710185 hasConceptScore W3085710185C41008148 @default.
- W3085710185 hasConceptScore W3085710185C50522688 @default.
- W3085710185 hasConceptScore W3085710185C50644808 @default.
- W3085710185 hasIssue "174" @default.
- W3085710185 hasLocation W30857101851 @default.
- W3085710185 hasOpenAccess W3085710185 @default.
- W3085710185 hasPrimaryLocation W30857101851 @default.
- W3085710185 hasRelatedWork W1542886316 @default.
- W3085710185 hasRelatedWork W1988115241 @default.
- W3085710185 hasRelatedWork W2103496339 @default.
- W3085710185 hasRelatedWork W2166116275 @default.
- W3085710185 hasRelatedWork W2528305538 @default.
- W3085710185 hasRelatedWork W2550848904 @default.
- W3085710185 hasRelatedWork W2785807865 @default.
- W3085710185 hasRelatedWork W2810699377 @default.
- W3085710185 hasRelatedWork W2914483840 @default.
- W3085710185 hasRelatedWork W2927724204 @default.
- W3085710185 hasRelatedWork W2946171505 @default.
- W3085710185 hasRelatedWork W2946302218 @default.
- W3085710185 hasRelatedWork W2956024278 @default.
- W3085710185 hasRelatedWork W2963146412 @default.
- W3085710185 hasRelatedWork W2965449430 @default.
- W3085710185 hasRelatedWork W2997330246 @default.
- W3085710185 hasRelatedWork W3031010353 @default.
- W3085710185 hasRelatedWork W3038705269 @default.
- W3085710185 hasRelatedWork W3084746729 @default.
- W3085710185 hasRelatedWork W3102511045 @default.
- W3085710185 hasVolume "21" @default.
- W3085710185 isParatext "false" @default.
- W3085710185 isRetracted "false" @default.
- W3085710185 magId "3085710185" @default.
- W3085710185 workType "article" @default.