Matches in SemOpenAlex for { <https://semopenalex.org/work/W2890267291> ?p ?o ?g. }
- W2890267291 endingPage "235" @default.
- W2890267291 startingPage "224" @default.
- W2890267291 abstract "This paper proposes a novel regularizer named Structured Decorrelation Constraint, to address both the generalization and optimization of deep neural networks, including multiple-layer perceptrons and convolutional neural networks. Our proposed regularizer reduces overfitting by breaking the co-adaptions between the neurons with an explicit penalty. As a result, the network is capable of learning non-redundant representations. Meanwhile, the proposed regularizer encourages the networks to learn structured high-level features to aid the networks’ optimization during training. To this end, neurons are constrained to behave obeying a group prior. Our regularizer applies to various types of layers, including fully connected layers, convolutional layers and normalization layers. The loss of our regularizer can be directly minimized along with the network’s classification loss by stochastic gradient descent. Experiments show that the proposed regularizer obviously relieves the overfitting problem of the existing deep networks. It yields much better performance on extensive datasets than the conventional regularizers like Dropout." @default.
- W2890267291 created "2018-09-27" @default.
- W2890267291 creator A5001529504 @default.
- W2890267291 creator A5038347628 @default.
- W2890267291 creator A5043887495 @default.
- W2890267291 creator A5061601210 @default.
- W2890267291 date "2019-02-01" @default.
- W2890267291 modified "2023-10-16" @default.
- W2890267291 title "Learning structured and non-redundant representations with deep neural networks" @default.
- W2890267291 cites W1961891967 @default.
- W2890267291 cites W2034978228 @default.
- W2890267291 cites W2086866337 @default.
- W2890267291 cites W2112796928 @default.
- W2890267291 cites W2155893237 @default.
- W2890267291 cites W2213612645 @default.
- W2890267291 cites W2291068538 @default.
- W2890267291 cites W2321627895 @default.
- W2890267291 cites W2331143823 @default.
- W2890267291 cites W2343514605 @default.
- W2890267291 cites W2412558220 @default.
- W2890267291 cites W2495808537 @default.
- W2890267291 cites W2507855991 @default.
- W2890267291 cites W2511374729 @default.
- W2890267291 cites W2552639984 @default.
- W2890267291 cites W2568998735 @default.
- W2890267291 cites W2600061660 @default.
- W2890267291 cites W2789677693 @default.
- W2890267291 cites W2964137095 @default.
- W2890267291 doi "https://doi.org/10.1016/j.patcog.2018.08.017" @default.
- W2890267291 hasPublicationYear "2019" @default.
- W2890267291 type Work @default.
- W2890267291 sameAs 2890267291 @default.
- W2890267291 citedByCount "17" @default.
- W2890267291 countsByYear W28902672912019 @default.
- W2890267291 countsByYear W28902672912020 @default.
- W2890267291 countsByYear W28902672912021 @default.
- W2890267291 countsByYear W28902672912022 @default.
- W2890267291 countsByYear W28902672912023 @default.
- W2890267291 crossrefType "journal-article" @default.
- W2890267291 hasAuthorship W2890267291A5001529504 @default.
- W2890267291 hasAuthorship W2890267291A5038347628 @default.
- W2890267291 hasAuthorship W2890267291A5043887495 @default.
- W2890267291 hasAuthorship W2890267291A5061601210 @default.
- W2890267291 hasConcept C108583219 @default.
- W2890267291 hasConcept C119857082 @default.
- W2890267291 hasConcept C134306372 @default.
- W2890267291 hasConcept C136886441 @default.
- W2890267291 hasConcept C144024400 @default.
- W2890267291 hasConcept C153180895 @default.
- W2890267291 hasConcept C154945302 @default.
- W2890267291 hasConcept C177148314 @default.
- W2890267291 hasConcept C19165224 @default.
- W2890267291 hasConcept C206688291 @default.
- W2890267291 hasConcept C22019652 @default.
- W2890267291 hasConcept C2524010 @default.
- W2890267291 hasConcept C2776036281 @default.
- W2890267291 hasConcept C2776145597 @default.
- W2890267291 hasConcept C2984842247 @default.
- W2890267291 hasConcept C33923547 @default.
- W2890267291 hasConcept C41008148 @default.
- W2890267291 hasConcept C50644808 @default.
- W2890267291 hasConcept C60908668 @default.
- W2890267291 hasConcept C81363708 @default.
- W2890267291 hasConceptScore W2890267291C108583219 @default.
- W2890267291 hasConceptScore W2890267291C119857082 @default.
- W2890267291 hasConceptScore W2890267291C134306372 @default.
- W2890267291 hasConceptScore W2890267291C136886441 @default.
- W2890267291 hasConceptScore W2890267291C144024400 @default.
- W2890267291 hasConceptScore W2890267291C153180895 @default.
- W2890267291 hasConceptScore W2890267291C154945302 @default.
- W2890267291 hasConceptScore W2890267291C177148314 @default.
- W2890267291 hasConceptScore W2890267291C19165224 @default.
- W2890267291 hasConceptScore W2890267291C206688291 @default.
- W2890267291 hasConceptScore W2890267291C22019652 @default.
- W2890267291 hasConceptScore W2890267291C2524010 @default.
- W2890267291 hasConceptScore W2890267291C2776036281 @default.
- W2890267291 hasConceptScore W2890267291C2776145597 @default.
- W2890267291 hasConceptScore W2890267291C2984842247 @default.
- W2890267291 hasConceptScore W2890267291C33923547 @default.
- W2890267291 hasConceptScore W2890267291C41008148 @default.
- W2890267291 hasConceptScore W2890267291C50644808 @default.
- W2890267291 hasConceptScore W2890267291C60908668 @default.
- W2890267291 hasConceptScore W2890267291C81363708 @default.
- W2890267291 hasLocation W28902672911 @default.
- W2890267291 hasOpenAccess W2890267291 @default.
- W2890267291 hasPrimaryLocation W28902672911 @default.
- W2890267291 hasRelatedWork W2781855238 @default.
- W2890267291 hasRelatedWork W2792147139 @default.
- W2890267291 hasRelatedWork W2897443685 @default.
- W2890267291 hasRelatedWork W2951100320 @default.
- W2890267291 hasRelatedWork W2998675825 @default.
- W2890267291 hasRelatedWork W3110700750 @default.
- W2890267291 hasRelatedWork W3128220493 @default.
- W2890267291 hasRelatedWork W4226354336 @default.
- W2890267291 hasRelatedWork W4298017035 @default.
- W2890267291 hasRelatedWork W4307654087 @default.
- W2890267291 hasVolume "86" @default.
- W2890267291 isParatext "false" @default.