Matches in SemOpenAlex for { <https://semopenalex.org/work/W2547432194> ?p ?o ?g. }
- W2547432194 endingPage "2054" @default.
- W2547432194 startingPage "2046" @default.
- W2547432194 abstract "We present a novel neural network algorithm, the Tensor Switching (TS) network, which generalizes the Rectified Linear Unit (ReLU) nonlinearity to tensor-valued hidden units. The TS network copies its entire input vector to different locations in an expanded representation, with the location determined by its hidden unit activity. In this way, even a simple linear readout from the TS representation can implement a highly expressive deep-network-like function. The TS network hence avoids the vanishing gradient problem by construction, at the cost of larger representation size. We develop several methods to train the TS network, including equivalent kernels for infinitely wide and deep TS networks, a one-pass linear learning algorithm, and two backpropagation-inspired representation learning algorithms. Our experimental results demonstrate that the TS network is indeed more expressive and consistently learns faster than standard ReLU networks." @default.
- W2547432194 created "2016-11-11" @default.
- W2547432194 creator A5008586286 @default.
- W2547432194 creator A5011428379 @default.
- W2547432194 creator A5052325705 @default.
- W2547432194 date "2016-12-05" @default.
- W2547432194 modified "2023-09-27" @default.
- W2547432194 title "Tensor switching networks" @default.
- W2547432194 cites W1026270304 @default.
- W2547432194 cites W1572063013 @default.
- W2547432194 cites W1665214252 @default.
- W2547432194 cites W1839868949 @default.
- W2547432194 cites W2076063813 @default.
- W2547432194 cites W2093231248 @default.
- W2547432194 cites W2097117768 @default.
- W2547432194 cites W2106662462 @default.
- W2547432194 cites W2130073327 @default.
- W2547432194 cites W2139622435 @default.
- W2547432194 cites W2289330286 @default.
- W2547432194 cites W2294059674 @default.
- W2547432194 cites W2302255633 @default.
- W2547432194 cites W2475171437 @default.
- W2547432194 cites W2511730936 @default.
- W2547432194 cites W2903628479 @default.
- W2547432194 cites W2951985269 @default.
- W2547432194 cites W2962835968 @default.
- W2547432194 cites W2963382180 @default.
- W2547432194 cites W2963504252 @default.
- W2547432194 cites W2964040467 @default.
- W2547432194 cites W2964243309 @default.
- W2547432194 cites W3101806332 @default.
- W2547432194 hasPublicationYear "2016" @default.
- W2547432194 type Work @default.
- W2547432194 sameAs 2547432194 @default.
- W2547432194 citedByCount "1" @default.
- W2547432194 countsByYear W25474321942019 @default.
- W2547432194 crossrefType "proceedings-article" @default.
- W2547432194 hasAuthorship W2547432194A5008586286 @default.
- W2547432194 hasAuthorship W2547432194A5011428379 @default.
- W2547432194 hasAuthorship W2547432194A5052325705 @default.
- W2547432194 hasConcept C108583219 @default.
- W2547432194 hasConcept C111472728 @default.
- W2547432194 hasConcept C11413529 @default.
- W2547432194 hasConcept C114614502 @default.
- W2547432194 hasConcept C138885662 @default.
- W2547432194 hasConcept C154945302 @default.
- W2547432194 hasConcept C155032097 @default.
- W2547432194 hasConcept C155281189 @default.
- W2547432194 hasConcept C17744445 @default.
- W2547432194 hasConcept C184720557 @default.
- W2547432194 hasConcept C199539241 @default.
- W2547432194 hasConcept C202444582 @default.
- W2547432194 hasConcept C2776359362 @default.
- W2547432194 hasConcept C2780586882 @default.
- W2547432194 hasConcept C33923547 @default.
- W2547432194 hasConcept C41008148 @default.
- W2547432194 hasConcept C50644808 @default.
- W2547432194 hasConcept C80444323 @default.
- W2547432194 hasConcept C94625758 @default.
- W2547432194 hasConceptScore W2547432194C108583219 @default.
- W2547432194 hasConceptScore W2547432194C111472728 @default.
- W2547432194 hasConceptScore W2547432194C11413529 @default.
- W2547432194 hasConceptScore W2547432194C114614502 @default.
- W2547432194 hasConceptScore W2547432194C138885662 @default.
- W2547432194 hasConceptScore W2547432194C154945302 @default.
- W2547432194 hasConceptScore W2547432194C155032097 @default.
- W2547432194 hasConceptScore W2547432194C155281189 @default.
- W2547432194 hasConceptScore W2547432194C17744445 @default.
- W2547432194 hasConceptScore W2547432194C184720557 @default.
- W2547432194 hasConceptScore W2547432194C199539241 @default.
- W2547432194 hasConceptScore W2547432194C202444582 @default.
- W2547432194 hasConceptScore W2547432194C2776359362 @default.
- W2547432194 hasConceptScore W2547432194C2780586882 @default.
- W2547432194 hasConceptScore W2547432194C33923547 @default.
- W2547432194 hasConceptScore W2547432194C41008148 @default.
- W2547432194 hasConceptScore W2547432194C50644808 @default.
- W2547432194 hasConceptScore W2547432194C80444323 @default.
- W2547432194 hasConceptScore W2547432194C94625758 @default.
- W2547432194 hasLocation W25474321941 @default.
- W2547432194 hasOpenAccess W2547432194 @default.
- W2547432194 hasPrimaryLocation W25474321941 @default.
- W2547432194 hasRelatedWork W1489744657 @default.
- W2547432194 hasRelatedWork W1544216806 @default.
- W2547432194 hasRelatedWork W2017557583 @default.
- W2547432194 hasRelatedWork W2063560698 @default.
- W2547432194 hasRelatedWork W2066915000 @default.
- W2547432194 hasRelatedWork W2114747792 @default.
- W2547432194 hasRelatedWork W2135302162 @default.
- W2547432194 hasRelatedWork W2151907713 @default.
- W2547432194 hasRelatedWork W2161388792 @default.
- W2547432194 hasRelatedWork W2181382924 @default.
- W2547432194 hasRelatedWork W2483809729 @default.
- W2547432194 hasRelatedWork W2490887468 @default.
- W2547432194 hasRelatedWork W2618540739 @default.
- W2547432194 hasRelatedWork W2790252067 @default.
- W2547432194 hasRelatedWork W2912908355 @default.
- W2547432194 hasRelatedWork W2952743641 @default.
- W2547432194 hasRelatedWork W2978030971 @default.