Matches in SemOpenAlex for { <https://semopenalex.org/work/W4308231892> ?p ?o ?g. }
Showing items 1 to 77 of
77
with 100 items per page.
- W4308231892 abstract "Abstract Artificial neural networks are widely used in data processing and Data Mining. In contrast to traditional artificial neural networks, deep neural networks employ many artificial neuron blocks and more than three layers. An increase in the number of neuron blocks and layers improves the approximation capabilities but leads to the effects of exploding and vanishing gradients. Deep neural networks often employ piece-wise activation functions like ReLU to overcome the effects of exploding and vanishing gradients. We propose F-neuron as an adaptive alternative to piece-wise activation functions. Like ReLU, F-neuron does not suffer from the effects of exploding and vanishing gradient. F-neuron changes its form during the training, can approximate any currently used function, and synthesize new task-specific functions. We show that F-neuron can synthesize new task-specific functions and achieve higher approximation quality in existing neural network architectures. We evaluate the performance of the F-neuron on two image classification datasets (Fashion-MNIST and CIFAR-10) in two different architectures (LeNet-5 and KerasNet)." @default.
- W4308231892 created "2022-11-09" @default.
- W4308231892 creator A5059642629 @default.
- W4308231892 creator A5076815371 @default.
- W4308231892 date "2022-11-03" @default.
- W4308231892 modified "2023-09-27" @default.
- W4308231892 title "Deep neural network based on F-neurons and its learning" @default.
- W4308231892 cites W1487043795 @default.
- W4308231892 cites W1505889263 @default.
- W4308231892 cites W1803890343 @default.
- W4308231892 cites W1988115241 @default.
- W4308231892 cites W2075594316 @default.
- W4308231892 cites W2076063813 @default.
- W4308231892 cites W2112796928 @default.
- W4308231892 cites W2156639609 @default.
- W4308231892 cites W2338768414 @default.
- W4308231892 cites W2919115771 @default.
- W4308231892 cites W4231804901 @default.
- W4308231892 cites W4231980566 @default.
- W4308231892 doi "https://doi.org/10.21203/rs.3.rs-2032768/v1" @default.
- W4308231892 hasPublicationYear "2022" @default.
- W4308231892 type Work @default.
- W4308231892 citedByCount "0" @default.
- W4308231892 crossrefType "posted-content" @default.
- W4308231892 hasAuthorship W4308231892A5059642629 @default.
- W4308231892 hasAuthorship W4308231892A5076815371 @default.
- W4308231892 hasBestOaLocation W43082318921 @default.
- W4308231892 hasConcept C108583219 @default.
- W4308231892 hasConcept C127413603 @default.
- W4308231892 hasConcept C14036430 @default.
- W4308231892 hasConcept C153180895 @default.
- W4308231892 hasConcept C154945302 @default.
- W4308231892 hasConcept C169760540 @default.
- W4308231892 hasConcept C190502265 @default.
- W4308231892 hasConcept C201995342 @default.
- W4308231892 hasConcept C2776990819 @default.
- W4308231892 hasConcept C2778794669 @default.
- W4308231892 hasConcept C2780451532 @default.
- W4308231892 hasConcept C2984842247 @default.
- W4308231892 hasConcept C38365724 @default.
- W4308231892 hasConcept C41008148 @default.
- W4308231892 hasConcept C50644808 @default.
- W4308231892 hasConcept C78458016 @default.
- W4308231892 hasConcept C86803240 @default.
- W4308231892 hasConceptScore W4308231892C108583219 @default.
- W4308231892 hasConceptScore W4308231892C127413603 @default.
- W4308231892 hasConceptScore W4308231892C14036430 @default.
- W4308231892 hasConceptScore W4308231892C153180895 @default.
- W4308231892 hasConceptScore W4308231892C154945302 @default.
- W4308231892 hasConceptScore W4308231892C169760540 @default.
- W4308231892 hasConceptScore W4308231892C190502265 @default.
- W4308231892 hasConceptScore W4308231892C201995342 @default.
- W4308231892 hasConceptScore W4308231892C2776990819 @default.
- W4308231892 hasConceptScore W4308231892C2778794669 @default.
- W4308231892 hasConceptScore W4308231892C2780451532 @default.
- W4308231892 hasConceptScore W4308231892C2984842247 @default.
- W4308231892 hasConceptScore W4308231892C38365724 @default.
- W4308231892 hasConceptScore W4308231892C41008148 @default.
- W4308231892 hasConceptScore W4308231892C50644808 @default.
- W4308231892 hasConceptScore W4308231892C78458016 @default.
- W4308231892 hasConceptScore W4308231892C86803240 @default.
- W4308231892 hasLocation W43082318921 @default.
- W4308231892 hasOpenAccess W4308231892 @default.
- W4308231892 hasPrimaryLocation W43082318921 @default.
- W4308231892 hasRelatedWork W2088610186 @default.
- W4308231892 hasRelatedWork W2550787675 @default.
- W4308231892 hasRelatedWork W2767072113 @default.
- W4308231892 hasRelatedWork W2773120646 @default.
- W4308231892 hasRelatedWork W2861440650 @default.
- W4308231892 hasRelatedWork W2888503998 @default.
- W4308231892 hasRelatedWork W2936783136 @default.
- W4308231892 hasRelatedWork W2947175736 @default.
- W4308231892 hasRelatedWork W3156786002 @default.
- W4308231892 hasRelatedWork W4308231892 @default.
- W4308231892 isParatext "false" @default.
- W4308231892 isRetracted "false" @default.
- W4308231892 workType "article" @default.