Matches in SemOpenAlex for { <https://semopenalex.org/work/W2021202584> ?p ?o ?g. }
Showing items 1 to 100 of
100
with 100 items per page.
- W2021202584 abstract "The Restricted Boltzmann Machine (RBM), a special case of general Boltzmann Machines and a typical Probabilistic Graphical Models, has attracted much attention in recent years due to its powerful ability in extracting features and representing the distribution underlying the training data. A most commonly used algorithm in learning RBMs is called Contrastive Divergence (CD) proposed by Hinton, which starts a Markov chain at a data point and runs the chain for only a few iterations to get a low variance estimator. However, when referring to a high-order RBM, since there are interactions among its visible layers, the gradient approximation via CD learning usually becomes far from the log-likelihood gradient and even may cause CD learning to fall into an infinite loop with high reconstruction error. In this paper, a new algorithm named Cyclic Contrastive Divergence (CCD) is introduced for learning high-order RBMs. Unlike the standard CD algorithm, CCD updates the parameters according to each visible layer in turn, by borrowing the idea of Cyclic Block Coordinate Descent method. To evaluate the performance of the proposed CCD algorithm, regarding to high-order RBMs learning, both algorithms CCD and standard CD are theoretically analyzed, including convergence, estimate upper bound and both biases comparison, from which the superiority of CCD learning is revealed. Experiments on MNIST dataset for the handwritten digit classification task are performed. The experimental results show that CCD is more applicable and consistently outperforms the standard CD in both convergent speed and performance." @default.
- W2021202584 created "2016-06-24" @default.
- W2021202584 creator A5004014436 @default.
- W2021202584 creator A5029667848 @default.
- W2021202584 creator A5067873183 @default.
- W2021202584 creator A5084685506 @default.
- W2021202584 date "2014-12-01" @default.
- W2021202584 modified "2023-10-16" @default.
- W2021202584 title "A Cyclic Contrastive Divergence Learning Algorithm for High-Order RBMs" @default.
- W2021202584 cites W1511867968 @default.
- W2021202584 cites W1540764732 @default.
- W2021202584 cites W1813659000 @default.
- W2021202584 cites W1964155876 @default.
- W2021202584 cites W1970819022 @default.
- W2021202584 cites W1974524469 @default.
- W2021202584 cites W2002016471 @default.
- W2021202584 cites W2020315425 @default.
- W2021202584 cites W2076878942 @default.
- W2021202584 cites W2097006534 @default.
- W2021202584 cites W2110361616 @default.
- W2021202584 cites W2115096495 @default.
- W2021202584 cites W2116064496 @default.
- W2021202584 cites W2156740722 @default.
- W2021202584 cites W2161000554 @default.
- W2021202584 cites W2179571827 @default.
- W2021202584 cites W2181867278 @default.
- W2021202584 cites W2203041758 @default.
- W2021202584 cites W2798766386 @default.
- W2021202584 cites W66838807 @default.
- W2021202584 doi "https://doi.org/10.1109/icmla.2014.18" @default.
- W2021202584 hasPublicationYear "2014" @default.
- W2021202584 type Work @default.
- W2021202584 sameAs 2021202584 @default.
- W2021202584 citedByCount "2" @default.
- W2021202584 countsByYear W20212025842015 @default.
- W2021202584 countsByYear W20212025842020 @default.
- W2021202584 crossrefType "proceedings-article" @default.
- W2021202584 hasAuthorship W2021202584A5004014436 @default.
- W2021202584 hasAuthorship W2021202584A5029667848 @default.
- W2021202584 hasAuthorship W2021202584A5067873183 @default.
- W2021202584 hasAuthorship W2021202584A5084685506 @default.
- W2021202584 hasConcept C105795698 @default.
- W2021202584 hasConcept C108583219 @default.
- W2021202584 hasConcept C11413529 @default.
- W2021202584 hasConcept C119857082 @default.
- W2021202584 hasConcept C138885662 @default.
- W2021202584 hasConcept C153180895 @default.
- W2021202584 hasConcept C154945302 @default.
- W2021202584 hasConcept C162324750 @default.
- W2021202584 hasConcept C185429906 @default.
- W2021202584 hasConcept C190502265 @default.
- W2021202584 hasConcept C192576344 @default.
- W2021202584 hasConcept C199354608 @default.
- W2021202584 hasConcept C207390915 @default.
- W2021202584 hasConcept C2524010 @default.
- W2021202584 hasConcept C2777210771 @default.
- W2021202584 hasConcept C2777303404 @default.
- W2021202584 hasConcept C33923547 @default.
- W2021202584 hasConcept C41008148 @default.
- W2021202584 hasConcept C41895202 @default.
- W2021202584 hasConcept C50522688 @default.
- W2021202584 hasConcept C98763669 @default.
- W2021202584 hasConceptScore W2021202584C105795698 @default.
- W2021202584 hasConceptScore W2021202584C108583219 @default.
- W2021202584 hasConceptScore W2021202584C11413529 @default.
- W2021202584 hasConceptScore W2021202584C119857082 @default.
- W2021202584 hasConceptScore W2021202584C138885662 @default.
- W2021202584 hasConceptScore W2021202584C153180895 @default.
- W2021202584 hasConceptScore W2021202584C154945302 @default.
- W2021202584 hasConceptScore W2021202584C162324750 @default.
- W2021202584 hasConceptScore W2021202584C185429906 @default.
- W2021202584 hasConceptScore W2021202584C190502265 @default.
- W2021202584 hasConceptScore W2021202584C192576344 @default.
- W2021202584 hasConceptScore W2021202584C199354608 @default.
- W2021202584 hasConceptScore W2021202584C207390915 @default.
- W2021202584 hasConceptScore W2021202584C2524010 @default.
- W2021202584 hasConceptScore W2021202584C2777210771 @default.
- W2021202584 hasConceptScore W2021202584C2777303404 @default.
- W2021202584 hasConceptScore W2021202584C33923547 @default.
- W2021202584 hasConceptScore W2021202584C41008148 @default.
- W2021202584 hasConceptScore W2021202584C41895202 @default.
- W2021202584 hasConceptScore W2021202584C50522688 @default.
- W2021202584 hasConceptScore W2021202584C98763669 @default.
- W2021202584 hasLocation W20212025841 @default.
- W2021202584 hasOpenAccess W2021202584 @default.
- W2021202584 hasPrimaryLocation W20212025841 @default.
- W2021202584 hasRelatedWork W1555941799 @default.
- W2021202584 hasRelatedWork W1900302034 @default.
- W2021202584 hasRelatedWork W2167735388 @default.
- W2021202584 hasRelatedWork W2187720631 @default.
- W2021202584 hasRelatedWork W2503569529 @default.
- W2021202584 hasRelatedWork W2892911634 @default.
- W2021202584 hasRelatedWork W2902142523 @default.
- W2021202584 hasRelatedWork W3082895349 @default.
- W2021202584 hasRelatedWork W3124769812 @default.
- W2021202584 hasRelatedWork W3156786002 @default.
- W2021202584 isParatext "false" @default.
- W2021202584 isRetracted "false" @default.
- W2021202584 magId "2021202584" @default.
- W2021202584 workType "article" @default.