FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2399926488> ?p ?o ?g. }

• W2399926488 abstract "Recurrent neural networks (RNNs) have drawn interest from machine learning researchers because of their effectiveness at preserving past inputs for time-varying data processing tasks. To understand the success and limitations of RNNs, it is critical that we advance our analysis of their fundamental memory properties. We focus on echo state networks (ESNs), which are RNNs with simple memoryless nodes and random connectivity. In most existing analyses, the short-term memory (STM) capacity results conclude that the ESN network size must scale linearly with the input size for unstructured inputs. The main contribution of this paper is to provide general results characterizing the STM capacity for linear ESNs with multidimensional input streams when the inputs have common low-dimensional structure: sparsity in a basis or significant statistical dependence between inputs. In both cases, we show that the number of nodes in the network must scale linearly with the information rate and poly-logarithmically with the ambient input dimension. The analysis relies on advanced applications of random matrix theory and results in explicit non-asymptotic bounds on the recovery error. Taken together, this analysis provides a significant step forward in our understanding of the STM properties in RNNs." @default.
• W2399926488 created "2016-06-24" @default.
• W2399926488 creator A5001139821 @default.
• W2399926488 creator A5011481913 @default.
• W2399926488 creator A5062177768 @default.
• W2399926488 date "2016-05-26" @default.
• W2399926488 modified "2023-09-27" @default.
• W2399926488 title "Distributed Sequence Memory of Multidimensional Inputs in Recurrent Networks" @default.
• W2399926488 cites W1128845248 @default.
• W2399926488 cites W1574851760 @default.
• W2399926488 cites W1732796048 @default.
• W2399926488 cites W1850742715 @default.
• W2399926488 cites W1947481528 @default.
• W2399926488 cites W1978845507 @default.
• W2399926488 cites W1989312953 @default.
• W2399926488 cites W1991564249 @default.
• W2399926488 cites W1991693778 @default.
• W2399926488 cites W1999174140 @default.
• W2399926488 cites W2002213454 @default.
• W2399926488 cites W2015861736 @default.
• W2399926488 cites W2016354087 @default.
• W2399926488 cites W2025054170 @default.
• W2399926488 cites W2026660464 @default.
• W2399926488 cites W2027802883 @default.
• W2399926488 cites W2029538739 @default.
• W2399926488 cites W2031327377 @default.
• W2399926488 cites W2047071281 @default.
• W2399926488 cites W2056380775 @default.
• W2399926488 cites W2059283452 @default.
• W2399926488 cites W2075510082 @default.
• W2399926488 cites W2089552695 @default.
• W2399926488 cites W2094045132 @default.
• W2399926488 cites W2100543212 @default.
• W2399926488 cites W2103179919 @default.
• W2399926488 cites W2103442863 @default.
• W2399926488 cites W2105398397 @default.
• W2399926488 cites W2109957347 @default.
• W2399926488 cites W2118525872 @default.
• W2399926488 cites W2118550318 @default.
• W2399926488 cites W2118706537 @default.
• W2399926488 cites W2119367754 @default.
• W2399926488 cites W2128904261 @default.
• W2399926488 cites W2130392029 @default.
• W2399926488 cites W2131037016 @default.
• W2399926488 cites W2134332047 @default.
• W2399926488 cites W2134603460 @default.
• W2399926488 cites W2137198385 @default.
• W2399926488 cites W2140867429 @default.
• W2399926488 cites W2141116650 @default.
• W2399926488 cites W2143612262 @default.
• W2399926488 cites W2145096794 @default.
• W2399926488 cites W2145889472 @default.
• W2399926488 cites W2148520247 @default.
• W2399926488 cites W2159213043 @default.
• W2399926488 cites W2160547390 @default.
• W2399926488 cites W2162451874 @default.
• W2399926488 cites W2168248818 @default.
• W2399926488 cites W2169488311 @default.
• W2399926488 cites W2170023599 @default.
• W2399926488 cites W2171865010 @default.
• W2399926488 cites W2177933918 @default.
• W2399926488 cites W2293634267 @default.
• W2399926488 cites W2339666411 @default.
• W2399926488 cites W2469594461 @default.
• W2399926488 cites W2504480743 @default.
• W2399926488 cites W2951008357 @default.
• W2399926488 cites W2953106257 @default.
• W2399926488 cites W2962699674 @default.
• W2399926488 cites W2962965465 @default.
• W2399926488 cites W2963078493 @default.
• W2399926488 cites W2963863227 @default.
• W2399926488 cites W2963934952 @default.
• W2399926488 hasPublicationYear "2016" @default.
• W2399926488 type Work @default.
• W2399926488 sameAs 2399926488 @default.
• W2399926488 citedByCount "0" @default.
• W2399926488 crossrefType "posted-content" @default.
• W2399926488 hasAuthorship W2399926488A5001139821 @default.
• W2399926488 hasAuthorship W2399926488A5011481913 @default.
• W2399926488 hasAuthorship W2399926488A5062177768 @default.
• W2399926488 hasConcept C111472728 @default.
• W2399926488 hasConcept C11413529 @default.
• W2399926488 hasConcept C120665830 @default.
• W2399926488 hasConcept C121332964 @default.
• W2399926488 hasConcept C12426560 @default.
• W2399926488 hasConcept C135796866 @default.
• W2399926488 hasConcept C138885662 @default.
• W2399926488 hasConcept C147168706 @default.
• W2399926488 hasConcept C154945302 @default.
• W2399926488 hasConcept C172025690 @default.
• W2399926488 hasConcept C192209626 @default.
• W2399926488 hasConcept C202444582 @default.
• W2399926488 hasConcept C2524010 @default.
• W2399926488 hasConcept C2778112365 @default.
• W2399926488 hasConcept C2778755073 @default.
• W2399926488 hasConcept C2780586882 @default.
• W2399926488 hasConcept C33676613 @default.
• W2399926488 hasConcept C33923547 @default.
• W2399926488 hasConcept C41008148 @default.
• W2399926488 hasConcept C50644808 @default.

• next