FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3173484311> ?p ?o ?g. }

• W3173484311 endingPage "326" @default.
• W3173484311 startingPage "314" @default.
• W3173484311 abstract "Restricted Boltzmann machines (RBMs) with a binary visible layer of size N and a Gaussian hidden layer of size P have been proved to be equivalent to a Hopfield neural network (HNN) made of N binary neurons and storing P patterns ξ, as long as the weights w in the former are identified with the patterns. Here we aim to leverage this equivalence to find effective initialisations for weights in the RBM when what is available is a set of noisy examples of each pattern, aiming to translate statistical mechanics background available for HNN to the study of RBM’s learning and retrieval abilities. In particular, given a set of definite, structureless patterns we build a sample of blurred examples and prove that the initialisation where w corresponds to the empirical average ξ¯ over the sample is a fixed point under stochastic gradient descent. Further, as a toy application of the duality between HNN and RBM, we consider the simplest random auto-encoder (a three layer network made of two RBMs coupled by their hidden layer) and evidence that, as long as the parameter setting corresponds to the retrieval region of the dual HNN, reconstruction and denoising can be accomplished trivially, while when the system is in the spin-glass phase inference algorithms are necessary. This questions the need for larger retrieval regions which we obtain by applying a Gram–Schmidt orthogonalisation to the patterns: in fact, this procedure yields to a set of patterns devoid of correlations and for which the largest retrieval region can be accomplished. Finally we consider an application of duality also in a structured case: we test this approach on the MNIST dataset, and obtain that the network performs already ∼67% of successful classifications, suggesting it can be exploited as a computationally-cheap pre-training." @default.
• W3173484311 created "2021-07-05" @default.
• W3173484311 creator A5005761274 @default.
• W3173484311 creator A5011672584 @default.
• W3173484311 creator A5041121941 @default.
• W3173484311 creator A5041457930 @default.
• W3173484311 date "2021-11-01" @default.
• W3173484311 modified "2023-10-16" @default.
• W3173484311 title "On the effective initialisation for restricted Boltzmann machines via duality with Hopfield model" @default.
• W3173484311 cites W1966240450 @default.
• W3173484311 cites W1995114494 @default.
• W3173484311 cites W2030358500 @default.
• W3173484311 cites W2042492924 @default.
• W3173484311 cites W2069129925 @default.
• W3173484311 cites W2116064496 @default.
• W3173484311 cites W2124537004 @default.
• W3173484311 cites W2125150881 @default.
• W3173484311 cites W2128084896 @default.
• W3173484311 cites W2136922672 @default.
• W3173484311 cites W2141442470 @default.
• W3173484311 cites W2490183153 @default.
• W3173484311 cites W2556341799 @default.
• W3173484311 cites W2567167069 @default.
• W3173484311 cites W2593807178 @default.
• W3173484311 cites W2919115771 @default.
• W3173484311 cites W2963208784 @default.
• W3173484311 cites W3030087702 @default.
• W3173484311 cites W3097007160 @default.
• W3173484311 cites W3098745735 @default.
• W3173484311 cites W3116923736 @default.
• W3173484311 cites W4376453414 @default.
• W3173484311 cites W3147726452 @default.
• W3173484311 doi "https://doi.org/10.1016/j.neunet.2021.06.017" @default.
• W3173484311 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/34175807" @default.
• W3173484311 hasPublicationYear "2021" @default.
• W3173484311 type Work @default.
• W3173484311 sameAs 3173484311 @default.
• W3173484311 citedByCount "9" @default.
• W3173484311 countsByYear W31734843112021 @default.
• W3173484311 countsByYear W31734843112022 @default.
• W3173484311 countsByYear W31734843112023 @default.
• W3173484311 crossrefType "journal-article" @default.
• W3173484311 hasAuthorship W3173484311A5005761274 @default.
• W3173484311 hasAuthorship W3173484311A5011672584 @default.
• W3173484311 hasAuthorship W3173484311A5041121941 @default.
• W3173484311 hasAuthorship W3173484311A5041457930 @default.
• W3173484311 hasConcept C11413529 @default.
• W3173484311 hasConcept C118615104 @default.
• W3173484311 hasConcept C121332964 @default.
• W3173484311 hasConcept C153083717 @default.
• W3173484311 hasConcept C153180895 @default.
• W3173484311 hasConcept C153258448 @default.
• W3173484311 hasConcept C154945302 @default.
• W3173484311 hasConcept C163716315 @default.
• W3173484311 hasConcept C177264268 @default.
• W3173484311 hasConcept C192576344 @default.
• W3173484311 hasConcept C199354608 @default.
• W3173484311 hasConcept C199360897 @default.
• W3173484311 hasConcept C206688291 @default.
• W3173484311 hasConcept C2776214188 @default.
• W3173484311 hasConcept C2780069185 @default.
• W3173484311 hasConcept C33923547 @default.
• W3173484311 hasConcept C41008148 @default.
• W3173484311 hasConcept C46421273 @default.
• W3173484311 hasConcept C48372109 @default.
• W3173484311 hasConcept C50644808 @default.
• W3173484311 hasConcept C62520636 @default.
• W3173484311 hasConcept C94375191 @default.
• W3173484311 hasConcept C97385483 @default.
• W3173484311 hasConceptScore W3173484311C11413529 @default.
• W3173484311 hasConceptScore W3173484311C118615104 @default.
• W3173484311 hasConceptScore W3173484311C121332964 @default.
• W3173484311 hasConceptScore W3173484311C153083717 @default.
• W3173484311 hasConceptScore W3173484311C153180895 @default.
• W3173484311 hasConceptScore W3173484311C153258448 @default.
• W3173484311 hasConceptScore W3173484311C154945302 @default.
• W3173484311 hasConceptScore W3173484311C163716315 @default.
• W3173484311 hasConceptScore W3173484311C177264268 @default.
• W3173484311 hasConceptScore W3173484311C192576344 @default.
• W3173484311 hasConceptScore W3173484311C199354608 @default.
• W3173484311 hasConceptScore W3173484311C199360897 @default.
• W3173484311 hasConceptScore W3173484311C206688291 @default.
• W3173484311 hasConceptScore W3173484311C2776214188 @default.
• W3173484311 hasConceptScore W3173484311C2780069185 @default.
• W3173484311 hasConceptScore W3173484311C33923547 @default.
• W3173484311 hasConceptScore W3173484311C41008148 @default.
• W3173484311 hasConceptScore W3173484311C46421273 @default.
• W3173484311 hasConceptScore W3173484311C48372109 @default.
• W3173484311 hasConceptScore W3173484311C50644808 @default.
• W3173484311 hasConceptScore W3173484311C62520636 @default.
• W3173484311 hasConceptScore W3173484311C94375191 @default.
• W3173484311 hasConceptScore W3173484311C97385483 @default.
• W3173484311 hasFunder F4320322510 @default.
• W3173484311 hasLocation W31734843111 @default.
• W3173484311 hasOpenAccess W3173484311 @default.
• W3173484311 hasPrimaryLocation W31734843111 @default.
• W3173484311 hasRelatedWork W177847060 @default.
• W3173484311 hasRelatedWork W2320963147 @default.

• next