Matches in SemOpenAlex for { <https://semopenalex.org/work/W4288019369> ?p ?o ?g. }
Showing items 1 to 59 of
59
with 100 items per page.
- W4288019369 abstract "This paper considers the growth in the length of one-dimensional trajectories as they are passed through deep ReLU neural networks, which, among other things, is one measure of the expressivity of deep networks. We generalise existing results, providing an alternative, simpler method for lower bounding expected trajectory growth through random networks, for a more general class of weights distributions, including sparsely connected networks. We illustrate this approach by deriving bounds for sparse-Gaussian, sparse-uniform, and sparse-discrete-valued random nets. We prove that trajectory growth can remain exponential in depth with these new distributions, including their sparse variants, with the sparsity parameter appearing in the base of the exponent." @default.
- W4288019369 created "2022-07-26" @default.
- W4288019369 creator A5059562923 @default.
- W4288019369 creator A5062089651 @default.
- W4288019369 date "2019-11-24" @default.
- W4288019369 modified "2023-09-23" @default.
- W4288019369 title "Trajectory growth lower bounds for random sparse deep ReLU networks" @default.
- W4288019369 doi "https://doi.org/10.48550/arxiv.1911.10651" @default.
- W4288019369 hasPublicationYear "2019" @default.
- W4288019369 type Work @default.
- W4288019369 citedByCount "0" @default.
- W4288019369 crossrefType "posted-content" @default.
- W4288019369 hasAuthorship W4288019369A5059562923 @default.
- W4288019369 hasAuthorship W4288019369A5062089651 @default.
- W4288019369 hasBestOaLocation W42880193691 @default.
- W4288019369 hasConcept C11413529 @default.
- W4288019369 hasConcept C121332964 @default.
- W4288019369 hasConcept C1276947 @default.
- W4288019369 hasConcept C13662910 @default.
- W4288019369 hasConcept C138885662 @default.
- W4288019369 hasConcept C154945302 @default.
- W4288019369 hasConcept C163716315 @default.
- W4288019369 hasConcept C2777212361 @default.
- W4288019369 hasConcept C2780388253 @default.
- W4288019369 hasConcept C33923547 @default.
- W4288019369 hasConcept C41008148 @default.
- W4288019369 hasConcept C41895202 @default.
- W4288019369 hasConcept C62520636 @default.
- W4288019369 hasConcept C63584917 @default.
- W4288019369 hasConceptScore W4288019369C11413529 @default.
- W4288019369 hasConceptScore W4288019369C121332964 @default.
- W4288019369 hasConceptScore W4288019369C1276947 @default.
- W4288019369 hasConceptScore W4288019369C13662910 @default.
- W4288019369 hasConceptScore W4288019369C138885662 @default.
- W4288019369 hasConceptScore W4288019369C154945302 @default.
- W4288019369 hasConceptScore W4288019369C163716315 @default.
- W4288019369 hasConceptScore W4288019369C2777212361 @default.
- W4288019369 hasConceptScore W4288019369C2780388253 @default.
- W4288019369 hasConceptScore W4288019369C33923547 @default.
- W4288019369 hasConceptScore W4288019369C41008148 @default.
- W4288019369 hasConceptScore W4288019369C41895202 @default.
- W4288019369 hasConceptScore W4288019369C62520636 @default.
- W4288019369 hasConceptScore W4288019369C63584917 @default.
- W4288019369 hasLocation W42880193691 @default.
- W4288019369 hasOpenAccess W4288019369 @default.
- W4288019369 hasPrimaryLocation W42880193691 @default.
- W4288019369 hasRelatedWork W1988914731 @default.
- W4288019369 hasRelatedWork W2008992764 @default.
- W4288019369 hasRelatedWork W2030128136 @default.
- W4288019369 hasRelatedWork W2041391279 @default.
- W4288019369 hasRelatedWork W2315492480 @default.
- W4288019369 hasRelatedWork W2381541036 @default.
- W4288019369 hasRelatedWork W2791133188 @default.
- W4288019369 hasRelatedWork W2990405359 @default.
- W4288019369 hasRelatedWork W3012966034 @default.
- W4288019369 hasRelatedWork W3107474891 @default.
- W4288019369 isParatext "false" @default.
- W4288019369 isRetracted "false" @default.
- W4288019369 workType "article" @default.