Matches in SemOpenAlex for { <https://semopenalex.org/work/W1652178237> ?p ?o ?g. }
Showing items 1 to 76 of
76
with 100 items per page.
- W1652178237 abstract "Neural network (NN) models have shown good performance on many image recognition benchmarks. Given large image datasets, these models typically have millions or billions of parameters that can easily lead to over-fitting without regularization. Dropout and DropConnect show their effectiveness of regularizing large fully connected layers within neural networks. In Dropout, each neural activation within the network is randomly set to zero with a probability during training. In DropConnect, a generalization of Dropout, each connection weight within the network is randomly set to zero with a probability instead. Both of the probabilities in Dropout and DropConnect are universal predefined constants. We propose Adaptive Local Drop (ALDrop), a novel regularization method that sets each connection weight within the network with a learned probability adaptive to the input image dataset using a locality-based measure. Experiments on several image recognition benchmarks show that our model outperforms Dropout and DropConnect." @default.
- W1652178237 created "2016-06-24" @default.
- W1652178237 creator A5011877804 @default.
- W1652178237 creator A5034054749 @default.
- W1652178237 creator A5055321661 @default.
- W1652178237 date "2015-07-01" @default.
- W1652178237 modified "2023-09-26" @default.
- W1652178237 title "Regularizing neural networks with adaptive local drop" @default.
- W1652178237 cites W132188046 @default.
- W1652178237 cites W1849277567 @default.
- W1652178237 cites W2112796928 @default.
- W1652178237 cites W2116064496 @default.
- W1652178237 cites W2134557905 @default.
- W1652178237 cites W4244577065 @default.
- W1652178237 doi "https://doi.org/10.1109/ijcnn.2015.7280567" @default.
- W1652178237 hasPublicationYear "2015" @default.
- W1652178237 type Work @default.
- W1652178237 sameAs 1652178237 @default.
- W1652178237 citedByCount "2" @default.
- W1652178237 countsByYear W16521782372017 @default.
- W1652178237 countsByYear W16521782372022 @default.
- W1652178237 crossrefType "proceedings-article" @default.
- W1652178237 hasAuthorship W1652178237A5011877804 @default.
- W1652178237 hasAuthorship W1652178237A5034054749 @default.
- W1652178237 hasAuthorship W1652178237A5055321661 @default.
- W1652178237 hasConcept C11413529 @default.
- W1652178237 hasConcept C115961682 @default.
- W1652178237 hasConcept C119857082 @default.
- W1652178237 hasConcept C138885662 @default.
- W1652178237 hasConcept C153180895 @default.
- W1652178237 hasConcept C154945302 @default.
- W1652178237 hasConcept C2776135515 @default.
- W1652178237 hasConcept C2776145597 @default.
- W1652178237 hasConcept C2779808786 @default.
- W1652178237 hasConcept C41008148 @default.
- W1652178237 hasConcept C41895202 @default.
- W1652178237 hasConcept C50644808 @default.
- W1652178237 hasConceptScore W1652178237C11413529 @default.
- W1652178237 hasConceptScore W1652178237C115961682 @default.
- W1652178237 hasConceptScore W1652178237C119857082 @default.
- W1652178237 hasConceptScore W1652178237C138885662 @default.
- W1652178237 hasConceptScore W1652178237C153180895 @default.
- W1652178237 hasConceptScore W1652178237C154945302 @default.
- W1652178237 hasConceptScore W1652178237C2776135515 @default.
- W1652178237 hasConceptScore W1652178237C2776145597 @default.
- W1652178237 hasConceptScore W1652178237C2779808786 @default.
- W1652178237 hasConceptScore W1652178237C41008148 @default.
- W1652178237 hasConceptScore W1652178237C41895202 @default.
- W1652178237 hasConceptScore W1652178237C50644808 @default.
- W1652178237 hasLocation W16521782371 @default.
- W1652178237 hasOpenAccess W1652178237 @default.
- W1652178237 hasPrimaryLocation W16521782371 @default.
- W1652178237 hasRelatedWork W2128011066 @default.
- W1652178237 hasRelatedWork W215155355 @default.
- W1652178237 hasRelatedWork W2593484309 @default.
- W1652178237 hasRelatedWork W2597357081 @default.
- W1652178237 hasRelatedWork W2785668315 @default.
- W1652178237 hasRelatedWork W2795416201 @default.
- W1652178237 hasRelatedWork W2913407569 @default.
- W1652178237 hasRelatedWork W2914317573 @default.
- W1652178237 hasRelatedWork W2922944537 @default.
- W1652178237 hasRelatedWork W2952316226 @default.
- W1652178237 hasRelatedWork W2962885042 @default.
- W1652178237 hasRelatedWork W2963704822 @default.
- W1652178237 hasRelatedWork W2984938990 @default.
- W1652178237 hasRelatedWork W2989079018 @default.
- W1652178237 hasRelatedWork W3007623470 @default.
- W1652178237 hasRelatedWork W3021623424 @default.
- W1652178237 hasRelatedWork W3089835766 @default.
- W1652178237 hasRelatedWork W3194674358 @default.
- W1652178237 hasRelatedWork W3204818005 @default.
- W1652178237 hasRelatedWork W2926362642 @default.
- W1652178237 isParatext "false" @default.
- W1652178237 isRetracted "false" @default.
- W1652178237 magId "1652178237" @default.
- W1652178237 workType "article" @default.