Matches in SemOpenAlex for { <https://semopenalex.org/work/W4320350802> ?p ?o ?g. }
Showing items 1 to 69 of
69
with 100 items per page.
- W4320350802 abstract "Various powerful deep neural network architectures have made great contribution to the exciting successes of deep learning in the past two decades. Among them, deep Residual Networks (ResNets) are of particular importance because they demonstrated great usefulness in computer vision by winning the first place in many deep learning competitions. Also, ResNets were the first class of neural networks in the development history of deep learning that are really deep. It is of mathematical interest and practical meaning to understand the convergence of deep ResNets. We aim at characterizing the convergence of deep ResNets as the depth tends to infinity in terms of the parameters of the networks. Toward this purpose, we first give a matrix-vector description of general deep neural networks with shortcut connections and formulate an explicit expression for the networks by using the notions of activation domains and activation matrices. The convergence is then reduced to the convergence of two series involving infinite products of non-square matrices. By studying the two series, we establish a sufficient condition for pointwise convergence of ResNets. Our result is able to give justification for the design of ResNets. We also conduct experiments on benchmark machine learning data to verify our results." @default.
- W4320350802 created "2023-02-13" @default.
- W4320350802 creator A5004379619 @default.
- W4320350802 creator A5023682610 @default.
- W4320350802 date "2022-05-13" @default.
- W4320350802 modified "2023-09-29" @default.
- W4320350802 title "Convergence Analysis of Deep Residual Networks" @default.
- W4320350802 doi "https://doi.org/10.48550/arxiv.2205.06571" @default.
- W4320350802 hasPublicationYear "2022" @default.
- W4320350802 type Work @default.
- W4320350802 citedByCount "0" @default.
- W4320350802 crossrefType "posted-content" @default.
- W4320350802 hasAuthorship W4320350802A5004379619 @default.
- W4320350802 hasAuthorship W4320350802A5023682610 @default.
- W4320350802 hasBestOaLocation W43203508021 @default.
- W4320350802 hasConcept C108583219 @default.
- W4320350802 hasConcept C11413529 @default.
- W4320350802 hasConcept C13280743 @default.
- W4320350802 hasConcept C134306372 @default.
- W4320350802 hasConcept C154945302 @default.
- W4320350802 hasConcept C155512373 @default.
- W4320350802 hasConcept C162324750 @default.
- W4320350802 hasConcept C185798385 @default.
- W4320350802 hasConcept C205649164 @default.
- W4320350802 hasConcept C27156116 @default.
- W4320350802 hasConcept C2777303404 @default.
- W4320350802 hasConcept C2777894999 @default.
- W4320350802 hasConcept C2777984123 @default.
- W4320350802 hasConcept C2984842247 @default.
- W4320350802 hasConcept C33923547 @default.
- W4320350802 hasConcept C38652104 @default.
- W4320350802 hasConcept C41008148 @default.
- W4320350802 hasConcept C50522688 @default.
- W4320350802 hasConcept C50644808 @default.
- W4320350802 hasConceptScore W4320350802C108583219 @default.
- W4320350802 hasConceptScore W4320350802C11413529 @default.
- W4320350802 hasConceptScore W4320350802C13280743 @default.
- W4320350802 hasConceptScore W4320350802C134306372 @default.
- W4320350802 hasConceptScore W4320350802C154945302 @default.
- W4320350802 hasConceptScore W4320350802C155512373 @default.
- W4320350802 hasConceptScore W4320350802C162324750 @default.
- W4320350802 hasConceptScore W4320350802C185798385 @default.
- W4320350802 hasConceptScore W4320350802C205649164 @default.
- W4320350802 hasConceptScore W4320350802C27156116 @default.
- W4320350802 hasConceptScore W4320350802C2777303404 @default.
- W4320350802 hasConceptScore W4320350802C2777894999 @default.
- W4320350802 hasConceptScore W4320350802C2777984123 @default.
- W4320350802 hasConceptScore W4320350802C2984842247 @default.
- W4320350802 hasConceptScore W4320350802C33923547 @default.
- W4320350802 hasConceptScore W4320350802C38652104 @default.
- W4320350802 hasConceptScore W4320350802C41008148 @default.
- W4320350802 hasConceptScore W4320350802C50522688 @default.
- W4320350802 hasConceptScore W4320350802C50644808 @default.
- W4320350802 hasLocation W43203508021 @default.
- W4320350802 hasOpenAccess W4320350802 @default.
- W4320350802 hasPrimaryLocation W43203508021 @default.
- W4320350802 hasRelatedWork W2014434717 @default.
- W4320350802 hasRelatedWork W2157943005 @default.
- W4320350802 hasRelatedWork W2164692126 @default.
- W4320350802 hasRelatedWork W2555447031 @default.
- W4320350802 hasRelatedWork W2790244305 @default.
- W4320350802 hasRelatedWork W2995787120 @default.
- W4320350802 hasRelatedWork W3202533293 @default.
- W4320350802 hasRelatedWork W4285682802 @default.
- W4320350802 hasRelatedWork W4302043355 @default.
- W4320350802 hasRelatedWork W4312660813 @default.
- W4320350802 isParatext "false" @default.
- W4320350802 isRetracted "false" @default.
- W4320350802 workType "article" @default.