FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3009390838> ?p ?o ?g. }

• W3009390838 abstract "We study the expressivity of deep neural networks. Measuring a network's complexity by its number of or by its number of neurons, we consider the class of functions for which the error of best approximation with networks of a given complexity decays at a certain rate when increasing the complexity budget. Using results from classical approximation theory, we show that this class can be endowed with a (quasi)-norm that makes it a linear function space, called approximation space. We establish that allowing the networks to have certain types of skip connections does not change the resulting approximation spaces. We also discuss the role of the network's nonlinearity (also known as activation function) on the resulting spaces, as well as the role of depth. For the popular ReLU nonlinearity and its powers, we relate the newly constructed spaces to classical Besov spaces. The established embeddings highlight that some functions of very low Besov smoothness can nevertheless be well approximated by neural networks, if these networks are sufficiently deep." @default.
• W3009390838 created "2020-03-13" @default.
• W3009390838 creator A5032519700 @default.
• W3009390838 creator A5041230403 @default.
• W3009390838 creator A5065334144 @default.
• W3009390838 creator A5090767423 @default.
• W3009390838 date "2019-05-03" @default.
• W3009390838 modified "2023-09-23" @default.
• W3009390838 title "Approximation spaces of deep neural networks" @default.
• W3009390838 cites W143004564 @default.
• W3009390838 cites W1490869886 @default.
• W3009390838 cites W1558239788 @default.
• W3009390838 cites W1582706191 @default.
• W3009390838 cites W1593542532 @default.
• W3009390838 cites W1677182931 @default.
• W3009390838 cites W1901129140 @default.
• W3009390838 cites W1963770618 @default.
• W3009390838 cites W196427885 @default.
• W3009390838 cites W1964402093 @default.
• W3009390838 cites W1971735090 @default.
• W3009390838 cites W1973631422 @default.
• W3009390838 cites W1988115241 @default.
• W3009390838 cites W1995341919 @default.
• W3009390838 cites W2002023932 @default.
• W3009390838 cites W2003410902 @default.
• W3009390838 cites W2006240266 @default.
• W3009390838 cites W2017290750 @default.
• W3009390838 cites W2020038074 @default.
• W3009390838 cites W2021676252 @default.
• W3009390838 cites W2022808402 @default.
• W3009390838 cites W2039329577 @default.
• W3009390838 cites W2073492708 @default.
• W3009390838 cites W2079224763 @default.
• W3009390838 cites W2103496339 @default.
• W3009390838 cites W2118103795 @default.
• W3009390838 cites W2126001294 @default.
• W3009390838 cites W2152593035 @default.
• W3009390838 cites W2158581396 @default.
• W3009390838 cites W2163605009 @default.
• W3009390838 cites W2166116275 @default.
• W3009390838 cites W2194775991 @default.
• W3009390838 cites W2274086169 @default.
• W3009390838 cites W2464921911 @default.
• W3009390838 cites W2513671774 @default.
• W3009390838 cites W2527189750 @default.
• W3009390838 cites W2528305538 @default.
• W3009390838 cites W2558580397 @default.
• W3009390838 cites W2623790698 @default.
• W3009390838 cites W2787302432 @default.
• W3009390838 cites W2800415562 @default.
• W3009390838 cites W2805915852 @default.
• W3009390838 cites W2919115771 @default.
• W3009390838 cites W2962742960 @default.
• W3009390838 cites W2962761333 @default.
• W3009390838 cites W2962845550 @default.
• W3009390838 cites W2962990163 @default.
• W3009390838 cites W2963146412 @default.
• W3009390838 cites W2964290344 @default.
• W3009390838 cites W3038830718 @default.
• W3009390838 cites W3099423575 @default.
• W3009390838 cites W3101765447 @default.
• W3009390838 cites W3122499440 @default.
• W3009390838 cites W3125537303 @default.
• W3009390838 cites W3126588943 @default.
• W3009390838 cites W3146803896 @default.
• W3009390838 cites W563287810 @default.
• W3009390838 cites W609840395 @default.
• W3009390838 cites W965573281 @default.
• W3009390838 cites W3200817088 @default.
• W3009390838 hasPublicationYear "2019" @default.
• W3009390838 type Work @default.
• W3009390838 sameAs 3009390838 @default.
• W3009390838 citedByCount "20" @default.
• W3009390838 countsByYear W30093908382018 @default.
• W3009390838 countsByYear W30093908382019 @default.
• W3009390838 countsByYear W30093908382020 @default.
• W3009390838 countsByYear W30093908382021 @default.
• W3009390838 countsByYear W30093908382022 @default.
• W3009390838 crossrefType "posted-content" @default.
• W3009390838 hasAuthorship W3009390838A5032519700 @default.
• W3009390838 hasAuthorship W3009390838A5041230403 @default.
• W3009390838 hasAuthorship W3009390838A5065334144 @default.
• W3009390838 hasAuthorship W3009390838A5090767423 @default.
• W3009390838 hasConcept C102634674 @default.
• W3009390838 hasConcept C104317684 @default.
• W3009390838 hasConcept C111919701 @default.
• W3009390838 hasConcept C114614502 @default.
• W3009390838 hasConcept C121332964 @default.
• W3009390838 hasConcept C122383733 @default.
• W3009390838 hasConcept C129698243 @default.
• W3009390838 hasConcept C134306372 @default.
• W3009390838 hasConcept C14036430 @default.
• W3009390838 hasConcept C142730499 @default.
• W3009390838 hasConcept C154945302 @default.
• W3009390838 hasConcept C158622935 @default.
• W3009390838 hasConcept C17744445 @default.
• W3009390838 hasConcept C184720557 @default.
• W3009390838 hasConcept C185592680 @default.
• W3009390838 hasConcept C191795146 @default.
• W3009390838 hasConcept C199539241 @default.

• next