FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2953197984> ?p ?o ?g. }

• W2953197984 abstract "Complex-valued neural networks (CVNNs) have been shown to be powerful nonlinear approximators when the input data can be properly modeled in the complex domain. One of the major challenges in scaling up CVNNs in practice is the design of complex activation functions. Recently, we proposed a novel framework for learning these activation functions neuron-wise in a data-dependent fashion, based on a cheap one-dimensional kernel expansion and the idea of kernel activation functions (KAFs). In this paper we argue that, despite its flexibility, this framework is still limited in the class of functions that can be modeled in the complex domain. We leverage the idea of widely linear complex kernels to extend the formulation, allowing for a richer expressiveness without an increase in the number of adaptable parameters. We test the resulting model on a set of complex-valued image classification benchmarks. Experimental results show that the resulting CVNNs can achieve higher accuracy while at the same time converging faster." @default.
• W2953197984 created "2019-06-27" @default.
• W2953197984 creator A5019647783 @default.
• W2953197984 creator A5020583119 @default.
• W2953197984 creator A5022153057 @default.
• W2953197984 creator A5039283820 @default.
• W2953197984 date "2019-02-06" @default.
• W2953197984 modified "2023-10-03" @default.
• W2953197984 title "Widely Linear Kernels for Complex-Valued Kernel Activation Functions" @default.
• W2953197984 cites W2058399474 @default.
• W2953197984 cites W2111052925 @default.
• W2953197984 cites W2124348301 @default.
• W2953197984 cites W2128618097 @default.
• W2953197984 cites W2134563598 @default.
• W2953197984 cites W2135858797 @default.
• W2953197984 cites W2146611938 @default.
• W2953197984 cites W2154558852 @default.
• W2953197984 cites W2164177859 @default.
• W2953197984 cites W2290300449 @default.
• W2953197984 cites W2342590538 @default.
• W2953197984 cites W2543913312 @default.
• W2953197984 cites W2590796488 @default.
• W2953197984 cites W2618946976 @default.
• W2953197984 cites W2750384547 @default.
• W2953197984 cites W2791940667 @default.
• W2953197984 cites W2805679356 @default.
• W2953197984 cites W2890355438 @default.
• W2953197984 cites W2895966359 @default.
• W2953197984 cites W2962907613 @default.
• W2953197984 cites W2963042606 @default.
• W2953197984 cites W639989723 @default.
• W2953197984 cites W93167102 @default.
• W2953197984 hasPublicationYear "2019" @default.
• W2953197984 type Work @default.
• W2953197984 sameAs 2953197984 @default.
• W2953197984 citedByCount "0" @default.
• W2953197984 crossrefType "posted-content" @default.
• W2953197984 hasAuthorship W2953197984A5019647783 @default.
• W2953197984 hasAuthorship W2953197984A5020583119 @default.
• W2953197984 hasAuthorship W2953197984A5022153057 @default.
• W2953197984 hasAuthorship W2953197984A5039283820 @default.
• W2953197984 hasConcept C105795698 @default.
• W2953197984 hasConcept C11413529 @default.
• W2953197984 hasConcept C121332964 @default.
• W2953197984 hasConcept C134306372 @default.
• W2953197984 hasConcept C153083717 @default.
• W2953197984 hasConcept C154945302 @default.
• W2953197984 hasConcept C158622935 @default.
• W2953197984 hasConcept C177264268 @default.
• W2953197984 hasConcept C199360897 @default.
• W2953197984 hasConcept C202444582 @default.
• W2953197984 hasConcept C2524010 @default.
• W2953197984 hasConcept C2780598303 @default.
• W2953197984 hasConcept C33923547 @default.
• W2953197984 hasConcept C36503486 @default.
• W2953197984 hasConcept C38365724 @default.
• W2953197984 hasConcept C41008148 @default.
• W2953197984 hasConcept C47822265 @default.
• W2953197984 hasConcept C50644808 @default.
• W2953197984 hasConcept C62520636 @default.
• W2953197984 hasConcept C74193536 @default.
• W2953197984 hasConcept C99844830 @default.
• W2953197984 hasConceptScore W2953197984C105795698 @default.
• W2953197984 hasConceptScore W2953197984C11413529 @default.
• W2953197984 hasConceptScore W2953197984C121332964 @default.
• W2953197984 hasConceptScore W2953197984C134306372 @default.
• W2953197984 hasConceptScore W2953197984C153083717 @default.
• W2953197984 hasConceptScore W2953197984C154945302 @default.
• W2953197984 hasConceptScore W2953197984C158622935 @default.
• W2953197984 hasConceptScore W2953197984C177264268 @default.
• W2953197984 hasConceptScore W2953197984C199360897 @default.
• W2953197984 hasConceptScore W2953197984C202444582 @default.
• W2953197984 hasConceptScore W2953197984C2524010 @default.
• W2953197984 hasConceptScore W2953197984C2780598303 @default.
• W2953197984 hasConceptScore W2953197984C33923547 @default.
• W2953197984 hasConceptScore W2953197984C36503486 @default.
• W2953197984 hasConceptScore W2953197984C38365724 @default.
• W2953197984 hasConceptScore W2953197984C41008148 @default.
• W2953197984 hasConceptScore W2953197984C47822265 @default.
• W2953197984 hasConceptScore W2953197984C50644808 @default.
• W2953197984 hasConceptScore W2953197984C62520636 @default.
• W2953197984 hasConceptScore W2953197984C74193536 @default.
• W2953197984 hasConceptScore W2953197984C99844830 @default.
• W2953197984 hasLocation W29531979841 @default.
• W2953197984 hasOpenAccess W2953197984 @default.
• W2953197984 hasPrimaryLocation W29531979841 @default.
• W2953197984 hasRelatedWork W2514504677 @default.
• W2953197984 hasRelatedWork W2753604357 @default.
• W2953197984 hasRelatedWork W2788991260 @default.
• W2953197984 hasRelatedWork W2800708770 @default.
• W2953197984 hasRelatedWork W2803499264 @default.
• W2953197984 hasRelatedWork W2884938415 @default.
• W2953197984 hasRelatedWork W2895966359 @default.
• W2953197984 hasRelatedWork W2912413295 @default.
• W2953197984 hasRelatedWork W2923664173 @default.
• W2953197984 hasRelatedWork W2962849464 @default.
• W2953197984 hasRelatedWork W2962907613 @default.
• W2953197984 hasRelatedWork W2980674230 @default.
• W2953197984 hasRelatedWork W2988972308 @default.
• W2953197984 hasRelatedWork W3016538117 @default.

• next