Matches in SemOpenAlex for { <https://semopenalex.org/work/W2988675576> ?p ?o ?g. }
- W2988675576 abstract "One of the fundamental problems in supervised classification and in machine learning in general, is the modelling of non-parametric invariances that exist in data. Most prior art has focused on enforcing priors in the form of invariances to parametric nuisance transformations that are expected to be present in data. Learning non-parametric invariances directly from data remains an important open problem. In this paper, we introduce a new architectural layer for convolutional networks which is capable of learning general invariances from data itself. This layer can learn invariance to non-parametric transformations and interestingly, motivates and incorporates permanent random connectomes, thereby being called Permanent Random Connectome Non-Parametric Transformation Networks (PRC-NPTN). PRC-NPTN networks are initialized with random connections (not just weights) which are a small subset of the connections in a fully connected convolution layer. Importantly, these connections in PRC-NPTNs once initialized remain permanent throughout training and testing. Permanent random connectomes make these architectures loosely more biologically plausible than many other mainstream network architectures which require highly ordered structures. We motivate randomly initialized connections as a simple method to learn invariance from data itself while invoking invariance towards multiple nuisance transformations simultaneously. We find that these randomly initialized permanent connections have positive effects on generalization, outperform much larger ConvNet baselines and the recently proposed Non-Parametric Transformation Network (NPTN) on benchmarks that enforce learning invariances from the data itself." @default.
- W2988675576 created "2019-11-22" @default.
- W2988675576 creator A5015371785 @default.
- W2988675576 creator A5033930017 @default.
- W2988675576 creator A5057959136 @default.
- W2988675576 date "2019-11-13" @default.
- W2988675576 modified "2023-09-27" @default.
- W2988675576 title "Learning Non-Parametric Invariances from Data with Permanent Random Connectomes" @default.
- W2988675576 cites W1576579425 @default.
- W2988675576 cites W1650052828 @default.
- W2988675576 cites W1907282891 @default.
- W2988675576 cites W2008008116 @default.
- W2988675576 cites W2009130045 @default.
- W2988675576 cites W2030933387 @default.
- W2988675576 cites W2092505576 @default.
- W2988675576 cites W2095705004 @default.
- W2988675576 cites W2112074816 @default.
- W2988675576 cites W2112796928 @default.
- W2988675576 cites W2116914011 @default.
- W2988675576 cites W2134664762 @default.
- W2988675576 cites W2136026194 @default.
- W2988675576 cites W2138621090 @default.
- W2988675576 cites W2159110831 @default.
- W2988675576 cites W2167383966 @default.
- W2988675576 cites W2180894497 @default.
- W2988675576 cites W2259950157 @default.
- W2988675576 cites W2279221249 @default.
- W2988675576 cites W2428707905 @default.
- W2988675576 cites W2531409750 @default.
- W2988675576 cites W2549139847 @default.
- W2988675576 cites W2552737632 @default.
- W2988675576 cites W2576915720 @default.
- W2988675576 cites W2593658258 @default.
- W2988675576 cites W2804047946 @default.
- W2988675576 cites W2962686123 @default.
- W2988675576 cites W2963125010 @default.
- W2988675576 cites W2963423708 @default.
- W2988675576 cites W2964207486 @default.
- W2988675576 cites W2965969868 @default.
- W2988675576 cites W2981985696 @default.
- W2988675576 cites W4919037 @default.
- W2988675576 cites W603908379 @default.
- W2988675576 hasPublicationYear "2019" @default.
- W2988675576 type Work @default.
- W2988675576 sameAs 2988675576 @default.
- W2988675576 citedByCount "0" @default.
- W2988675576 crossrefType "posted-content" @default.
- W2988675576 hasAuthorship W2988675576A5015371785 @default.
- W2988675576 hasAuthorship W2988675576A5033930017 @default.
- W2988675576 hasAuthorship W2988675576A5057959136 @default.
- W2988675576 hasConcept C104317684 @default.
- W2988675576 hasConcept C105795698 @default.
- W2988675576 hasConcept C11413529 @default.
- W2988675576 hasConcept C117251300 @default.
- W2988675576 hasConcept C119857082 @default.
- W2988675576 hasConcept C134306372 @default.
- W2988675576 hasConcept C154945302 @default.
- W2988675576 hasConcept C177148314 @default.
- W2988675576 hasConcept C185592680 @default.
- W2988675576 hasConcept C204241405 @default.
- W2988675576 hasConcept C33923547 @default.
- W2988675576 hasConcept C41008148 @default.
- W2988675576 hasConcept C55493867 @default.
- W2988675576 hasConcept C80444323 @default.
- W2988675576 hasConceptScore W2988675576C104317684 @default.
- W2988675576 hasConceptScore W2988675576C105795698 @default.
- W2988675576 hasConceptScore W2988675576C11413529 @default.
- W2988675576 hasConceptScore W2988675576C117251300 @default.
- W2988675576 hasConceptScore W2988675576C119857082 @default.
- W2988675576 hasConceptScore W2988675576C134306372 @default.
- W2988675576 hasConceptScore W2988675576C154945302 @default.
- W2988675576 hasConceptScore W2988675576C177148314 @default.
- W2988675576 hasConceptScore W2988675576C185592680 @default.
- W2988675576 hasConceptScore W2988675576C204241405 @default.
- W2988675576 hasConceptScore W2988675576C33923547 @default.
- W2988675576 hasConceptScore W2988675576C41008148 @default.
- W2988675576 hasConceptScore W2988675576C55493867 @default.
- W2988675576 hasConceptScore W2988675576C80444323 @default.
- W2988675576 hasLocation W29886755761 @default.
- W2988675576 hasOpenAccess W2988675576 @default.
- W2988675576 hasPrimaryLocation W29886755761 @default.
- W2988675576 hasRelatedWork W147969812 @default.
- W2988675576 hasRelatedWork W1698663318 @default.
- W2988675576 hasRelatedWork W1960700909 @default.
- W2988675576 hasRelatedWork W2160660594 @default.
- W2988675576 hasRelatedWork W2208906386 @default.
- W2988675576 hasRelatedWork W2529448179 @default.
- W2988675576 hasRelatedWork W2803766319 @default.
- W2988675576 hasRelatedWork W2808987817 @default.
- W2988675576 hasRelatedWork W2904613663 @default.
- W2988675576 hasRelatedWork W2965969868 @default.
- W2988675576 hasRelatedWork W2979915069 @default.
- W2988675576 hasRelatedWork W2994486311 @default.
- W2988675576 hasRelatedWork W3036169520 @default.
- W2988675576 hasRelatedWork W3043057662 @default.
- W2988675576 hasRelatedWork W3129086097 @default.
- W2988675576 hasRelatedWork W3134779043 @default.
- W2988675576 hasRelatedWork W3166703191 @default.
- W2988675576 hasRelatedWork W3180050610 @default.
- W2988675576 hasRelatedWork W3181738625 @default.