Matches in SemOpenAlex for { <https://semopenalex.org/work/W4289492479> ?p ?o ?g. }
Showing items 1 to 77 of
77
with 100 items per page.
- W4289492479 abstract "Statistical characteristics of deep network representations, such as sparsity and correlation, are known to be relevant to the performance and interpretability of deep learning. When a statistical characteristic is desired, often an adequate regularizer can be designed and applied during the training phase. Typically, such a regularizer aims to manipulate a statistical characteristic over all classes together. For classification tasks, however, it might be advantageous to enforce the desired characteristic per class such that different classes can be better distinguished. Motivated by the idea, we design two class-wise regularizers that explicitly utilize class information: class-wise Covariance Regularizer (cw-CR) and class-wise Variance Regularizer (cw-VR). cw-CR targets to reduce the covariance of representations calculated from the same class samples for encouraging feature independence. cw-VR is similar, but variance instead of covariance is targeted to improve feature compactness. For the sake of completeness, their counterparts without using class information, Covariance Regularizer (CR) and Variance Regularizer (VR), are considered together. The four regularizers are conceptually simple and computationally very efficient, and the visualization shows that the regularizers indeed perform distinct representation shaping. In terms of classification performance, significant improvements over the baseline and L1/L2 weight regularization methods were found for 21 out of 22 tasks over popular benchmark datasets. In particular, cw-VR achieved the best performance for 13 tasks including ResNet-32/110." @default.
- W4289492479 created "2022-08-03" @default.
- W4289492479 creator A5056032525 @default.
- W4289492479 creator A5066676734 @default.
- W4289492479 date "2018-09-24" @default.
- W4289492479 modified "2023-09-30" @default.
- W4289492479 title "Utilizing Class Information for Deep Network Representation Shaping" @default.
- W4289492479 doi "https://doi.org/10.48550/arxiv.1809.09307" @default.
- W4289492479 hasPublicationYear "2018" @default.
- W4289492479 type Work @default.
- W4289492479 citedByCount "0" @default.
- W4289492479 crossrefType "posted-content" @default.
- W4289492479 hasAuthorship W4289492479A5056032525 @default.
- W4289492479 hasAuthorship W4289492479A5066676734 @default.
- W4289492479 hasBestOaLocation W42894924791 @default.
- W4289492479 hasConcept C105795698 @default.
- W4289492479 hasConcept C119857082 @default.
- W4289492479 hasConcept C121955636 @default.
- W4289492479 hasConcept C13280743 @default.
- W4289492479 hasConcept C138885662 @default.
- W4289492479 hasConcept C144133560 @default.
- W4289492479 hasConcept C153180895 @default.
- W4289492479 hasConcept C154945302 @default.
- W4289492479 hasConcept C17744445 @default.
- W4289492479 hasConcept C178650346 @default.
- W4289492479 hasConcept C185798385 @default.
- W4289492479 hasConcept C196083921 @default.
- W4289492479 hasConcept C199539241 @default.
- W4289492479 hasConcept C205649164 @default.
- W4289492479 hasConcept C2776135515 @default.
- W4289492479 hasConcept C2776359362 @default.
- W4289492479 hasConcept C2776401178 @default.
- W4289492479 hasConcept C2777212361 @default.
- W4289492479 hasConcept C2781067378 @default.
- W4289492479 hasConcept C33923547 @default.
- W4289492479 hasConcept C41008148 @default.
- W4289492479 hasConcept C41895202 @default.
- W4289492479 hasConcept C94625758 @default.
- W4289492479 hasConceptScore W4289492479C105795698 @default.
- W4289492479 hasConceptScore W4289492479C119857082 @default.
- W4289492479 hasConceptScore W4289492479C121955636 @default.
- W4289492479 hasConceptScore W4289492479C13280743 @default.
- W4289492479 hasConceptScore W4289492479C138885662 @default.
- W4289492479 hasConceptScore W4289492479C144133560 @default.
- W4289492479 hasConceptScore W4289492479C153180895 @default.
- W4289492479 hasConceptScore W4289492479C154945302 @default.
- W4289492479 hasConceptScore W4289492479C17744445 @default.
- W4289492479 hasConceptScore W4289492479C178650346 @default.
- W4289492479 hasConceptScore W4289492479C185798385 @default.
- W4289492479 hasConceptScore W4289492479C196083921 @default.
- W4289492479 hasConceptScore W4289492479C199539241 @default.
- W4289492479 hasConceptScore W4289492479C205649164 @default.
- W4289492479 hasConceptScore W4289492479C2776135515 @default.
- W4289492479 hasConceptScore W4289492479C2776359362 @default.
- W4289492479 hasConceptScore W4289492479C2776401178 @default.
- W4289492479 hasConceptScore W4289492479C2777212361 @default.
- W4289492479 hasConceptScore W4289492479C2781067378 @default.
- W4289492479 hasConceptScore W4289492479C33923547 @default.
- W4289492479 hasConceptScore W4289492479C41008148 @default.
- W4289492479 hasConceptScore W4289492479C41895202 @default.
- W4289492479 hasConceptScore W4289492479C94625758 @default.
- W4289492479 hasLocation W42894924791 @default.
- W4289492479 hasOpenAccess W4289492479 @default.
- W4289492479 hasPrimaryLocation W42894924791 @default.
- W4289492479 hasRelatedWork W2970447476 @default.
- W4289492479 hasRelatedWork W2984803561 @default.
- W4289492479 hasRelatedWork W3006943036 @default.
- W4289492479 hasRelatedWork W3012234327 @default.
- W4289492479 hasRelatedWork W3126014626 @default.
- W4289492479 hasRelatedWork W3191046242 @default.
- W4289492479 hasRelatedWork W3203961807 @default.
- W4289492479 hasRelatedWork W4206534706 @default.
- W4289492479 hasRelatedWork W4229079080 @default.
- W4289492479 hasRelatedWork W4294031299 @default.
- W4289492479 isParatext "false" @default.
- W4289492479 isRetracted "false" @default.
- W4289492479 workType "article" @default.