Matches in SemOpenAlex for { <https://semopenalex.org/work/W3137695714> ?p ?o ?g. }
Showing items 1 to 81 of
81
with 100 items per page.
- W3137695714 abstract "Despite their massive size, successful deep artificial neural networks can exhibit a remarkably small difference between training and test performance. Conventional wisdom attributes small generalization error either to properties of the model family, or to the regularization techniques used during training. Through extensive systematic experiments, we show how these traditional approaches fail to explain why large neural networks generalize well in practice. Specifically, our experiments establish that state-of-the-art convolutional networks for image classification trained with stochastic gradient methods easily fit a random labeling of the training data. This phenomenon is qualitatively unaffected by explicit regularization, and occurs even if we replace the true images by completely unstructured random noise. We corroborate these experimental findings with a theoretical construction showing that simple depth two neural networks already have perfect finite sample expressivity as soon as the number of parameters exceeds the number of data points as it usually does in practice. We interpret our experimental findings by comparison with traditional models." @default.
- W3137695714 created "2021-03-29" @default.
- W3137695714 creator A5003562101 @default.
- W3137695714 creator A5012870568 @default.
- W3137695714 creator A5016054994 @default.
- W3137695714 creator A5017529415 @default.
- W3137695714 creator A5039915143 @default.
- W3137695714 date "2016-11-10" @default.
- W3137695714 modified "2023-10-02" @default.
- W3137695714 title "Understanding deep learning requires rethinking generalization" @default.
- W3137695714 doi "https://doi.org/10.48550/arxiv.1611.03530" @default.
- W3137695714 hasPublicationYear "2016" @default.
- W3137695714 type Work @default.
- W3137695714 sameAs 3137695714 @default.
- W3137695714 citedByCount "1236" @default.
- W3137695714 countsByYear W31376957142016 @default.
- W3137695714 countsByYear W31376957142017 @default.
- W3137695714 countsByYear W31376957142018 @default.
- W3137695714 countsByYear W31376957142019 @default.
- W3137695714 countsByYear W31376957142020 @default.
- W3137695714 countsByYear W31376957142021 @default.
- W3137695714 countsByYear W31376957142022 @default.
- W3137695714 countsByYear W31376957142023 @default.
- W3137695714 crossrefType "posted-content" @default.
- W3137695714 hasAuthorship W3137695714A5003562101 @default.
- W3137695714 hasAuthorship W3137695714A5012870568 @default.
- W3137695714 hasAuthorship W3137695714A5016054994 @default.
- W3137695714 hasAuthorship W3137695714A5017529415 @default.
- W3137695714 hasAuthorship W3137695714A5039915143 @default.
- W3137695714 hasBestOaLocation W31376957141 @default.
- W3137695714 hasConcept C108583219 @default.
- W3137695714 hasConcept C111472728 @default.
- W3137695714 hasConcept C11413529 @default.
- W3137695714 hasConcept C117765406 @default.
- W3137695714 hasConcept C119857082 @default.
- W3137695714 hasConcept C134306372 @default.
- W3137695714 hasConcept C138885662 @default.
- W3137695714 hasConcept C154945302 @default.
- W3137695714 hasConcept C177148314 @default.
- W3137695714 hasConcept C2776135515 @default.
- W3137695714 hasConcept C2780586882 @default.
- W3137695714 hasConcept C2984842247 @default.
- W3137695714 hasConcept C33923547 @default.
- W3137695714 hasConcept C41008148 @default.
- W3137695714 hasConcept C50644808 @default.
- W3137695714 hasConcept C5465570 @default.
- W3137695714 hasConcept C81363708 @default.
- W3137695714 hasConceptScore W3137695714C108583219 @default.
- W3137695714 hasConceptScore W3137695714C111472728 @default.
- W3137695714 hasConceptScore W3137695714C11413529 @default.
- W3137695714 hasConceptScore W3137695714C117765406 @default.
- W3137695714 hasConceptScore W3137695714C119857082 @default.
- W3137695714 hasConceptScore W3137695714C134306372 @default.
- W3137695714 hasConceptScore W3137695714C138885662 @default.
- W3137695714 hasConceptScore W3137695714C154945302 @default.
- W3137695714 hasConceptScore W3137695714C177148314 @default.
- W3137695714 hasConceptScore W3137695714C2776135515 @default.
- W3137695714 hasConceptScore W3137695714C2780586882 @default.
- W3137695714 hasConceptScore W3137695714C2984842247 @default.
- W3137695714 hasConceptScore W3137695714C33923547 @default.
- W3137695714 hasConceptScore W3137695714C41008148 @default.
- W3137695714 hasConceptScore W3137695714C50644808 @default.
- W3137695714 hasConceptScore W3137695714C5465570 @default.
- W3137695714 hasConceptScore W3137695714C81363708 @default.
- W3137695714 hasLocation W31376957141 @default.
- W3137695714 hasOpenAccess W3137695714 @default.
- W3137695714 hasPrimaryLocation W31376957141 @default.
- W3137695714 hasRelatedWork W2279398222 @default.
- W3137695714 hasRelatedWork W2752860283 @default.
- W3137695714 hasRelatedWork W2900725142 @default.
- W3137695714 hasRelatedWork W2903815729 @default.
- W3137695714 hasRelatedWork W2906967080 @default.
- W3137695714 hasRelatedWork W2915754718 @default.
- W3137695714 hasRelatedWork W3133861977 @default.
- W3137695714 hasRelatedWork W4299822940 @default.
- W3137695714 hasRelatedWork W4311257506 @default.
- W3137695714 hasRelatedWork W4327499916 @default.
- W3137695714 isParatext "false" @default.
- W3137695714 isRetracted "false" @default.
- W3137695714 magId "3137695714" @default.
- W3137695714 workType "article" @default.