Matches in SemOpenAlex for { <https://semopenalex.org/work/W2963239103> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W2963239103 abstract "Recent works have cast some light on the mystery of why deep nets fit any data and generalize despite being very overparametrized. This paper analyzes training and generalization for a simple 2-layer ReLU net with random initialization, and provides the following improvements over recent works: (i) Using a tighter characterization of training speed than recent papers, an explanation for why training a neural net with random labels leads to slower training, as originally observed in [Zhang et al. ICLR'17]. (ii) Generalization bound independent of network size, using a data-dependent complexity measure. Our measure distinguishes clearly between random labels and true labels on MNIST and CIFAR, as shown by experiments. Moreover, recent papers require sample complexity to increase (slowly) with the size, while our sample complexity is completely independent of the network size. (iii) Learnability of a broad class of smooth functions by 2-layer ReLU nets trained via gradient descent. The key idea is to track dynamics of training and generalization via properties of a related kernel." @default.
- W2963239103 created "2019-07-30" @default.
- W2963239103 creator A5013194814 @default.
- W2963239103 creator A5033061754 @default.
- W2963239103 creator A5062378128 @default.
- W2963239103 creator A5062744744 @default.
- W2963239103 creator A5079951047 @default.
- W2963239103 date "2019-01-24" @default.
- W2963239103 modified "2023-10-18" @default.
- W2963239103 title "Fine-Grained Analysis of Optimization and Generalization for Overparameterized Two-Layer Neural Networks" @default.
- W2963239103 doi "https://doi.org/10.48550/arxiv.1901.08584" @default.
- W2963239103 hasPublicationYear "2019" @default.
- W2963239103 type Work @default.
- W2963239103 sameAs 2963239103 @default.
- W2963239103 citedByCount "167" @default.
- W2963239103 countsByYear W29632391032019 @default.
- W2963239103 countsByYear W29632391032020 @default.
- W2963239103 countsByYear W29632391032021 @default.
- W2963239103 countsByYear W29632391032022 @default.
- W2963239103 crossrefType "posted-content" @default.
- W2963239103 hasAuthorship W2963239103A5013194814 @default.
- W2963239103 hasAuthorship W2963239103A5033061754 @default.
- W2963239103 hasAuthorship W2963239103A5062378128 @default.
- W2963239103 hasAuthorship W2963239103A5062744744 @default.
- W2963239103 hasAuthorship W2963239103A5079951047 @default.
- W2963239103 hasBestOaLocation W29632391031 @default.
- W2963239103 hasConcept C111472728 @default.
- W2963239103 hasConcept C11413529 @default.
- W2963239103 hasConcept C114466953 @default.
- W2963239103 hasConcept C124101348 @default.
- W2963239103 hasConcept C134306372 @default.
- W2963239103 hasConcept C138885662 @default.
- W2963239103 hasConcept C153258448 @default.
- W2963239103 hasConcept C154945302 @default.
- W2963239103 hasConcept C177148314 @default.
- W2963239103 hasConcept C178790620 @default.
- W2963239103 hasConcept C185592680 @default.
- W2963239103 hasConcept C190502265 @default.
- W2963239103 hasConcept C199360897 @default.
- W2963239103 hasConcept C2777723229 @default.
- W2963239103 hasConcept C2779227376 @default.
- W2963239103 hasConcept C2780009758 @default.
- W2963239103 hasConcept C2780586882 @default.
- W2963239103 hasConcept C33923547 @default.
- W2963239103 hasConcept C41008148 @default.
- W2963239103 hasConcept C50644808 @default.
- W2963239103 hasConceptScore W2963239103C111472728 @default.
- W2963239103 hasConceptScore W2963239103C11413529 @default.
- W2963239103 hasConceptScore W2963239103C114466953 @default.
- W2963239103 hasConceptScore W2963239103C124101348 @default.
- W2963239103 hasConceptScore W2963239103C134306372 @default.
- W2963239103 hasConceptScore W2963239103C138885662 @default.
- W2963239103 hasConceptScore W2963239103C153258448 @default.
- W2963239103 hasConceptScore W2963239103C154945302 @default.
- W2963239103 hasConceptScore W2963239103C177148314 @default.
- W2963239103 hasConceptScore W2963239103C178790620 @default.
- W2963239103 hasConceptScore W2963239103C185592680 @default.
- W2963239103 hasConceptScore W2963239103C190502265 @default.
- W2963239103 hasConceptScore W2963239103C199360897 @default.
- W2963239103 hasConceptScore W2963239103C2777723229 @default.
- W2963239103 hasConceptScore W2963239103C2779227376 @default.
- W2963239103 hasConceptScore W2963239103C2780009758 @default.
- W2963239103 hasConceptScore W2963239103C2780586882 @default.
- W2963239103 hasConceptScore W2963239103C33923547 @default.
- W2963239103 hasConceptScore W2963239103C41008148 @default.
- W2963239103 hasConceptScore W2963239103C50644808 @default.
- W2963239103 hasLocation W29632391031 @default.
- W2963239103 hasOpenAccess W2963239103 @default.
- W2963239103 hasPrimaryLocation W29632391031 @default.
- W2963239103 hasRelatedWork W2368205053 @default.
- W2963239103 hasRelatedWork W2765525762 @default.
- W2963239103 hasRelatedWork W2887630004 @default.
- W2963239103 hasRelatedWork W2911867426 @default.
- W2963239103 hasRelatedWork W2963239103 @default.
- W2963239103 hasRelatedWork W2978031686 @default.
- W2963239103 hasRelatedWork W2996067004 @default.
- W2963239103 hasRelatedWork W3193949130 @default.
- W2963239103 hasRelatedWork W4206108497 @default.
- W2963239103 hasRelatedWork W4289700143 @default.
- W2963239103 isParatext "false" @default.
- W2963239103 isRetracted "false" @default.
- W2963239103 magId "2963239103" @default.
- W2963239103 workType "article" @default.