Matches in SemOpenAlex for { <https://semopenalex.org/work/W2099479976> ?p ?o ?g. }
- W2099479976 endingPage "1901" @default.
- W2099479976 startingPage "1849" @default.
- W2099479976 abstract "We propose a novel approach for nonlinear regression using a two-layer neural network (NN) model structure with sparsity-favoring hierarchical priors on the network weights. We present an expectation propagation (EP) approach for approximate integration over the posterior distribution of the weights, the hierarchical scale parameters of the priors, and the residual scale. Using a factorized posterior approximation we derive a computationally effcient algorithm, whose complexity scales similarly to an ensemble of independent sparse linear models. The approach enables flexible definition of weight priors with different sparseness properties such as independent Laplace priors with a common scale parameter or Gaussian automatic relevance determination (ARD) priors with different relevance parameters for all inputs. The approach can be extended beyond standard activation functions and NN model structures to form flexible nonlinear predictors from multiple sparse linear models. The effects of the hierarchical priors and the predictive performance of the algorithm are assessed using both simulated and real-world data. Comparisons are made to two alternative models with ARD priors: a Gaussian process with a NN covariance function and marginal maximum a posteriori estimates of the relevance parameters, and a NN with Markov chain Monte Carlo integration over all the unknown model parameters." @default.
- W2099479976 created "2016-06-24" @default.
- W2099479976 creator A5047269552 @default.
- W2099479976 creator A5048712208 @default.
- W2099479976 creator A5075462759 @default.
- W2099479976 date "2014-01-01" @default.
- W2099479976 modified "2023-09-25" @default.
- W2099479976 title "Expectation propagation for neural networks with sparsity-promoting priors" @default.
- W2099479976 cites W136174036 @default.
- W2099479976 cites W1515272691 @default.
- W2099479976 cites W1530235965 @default.
- W2099479976 cites W1567512734 @default.
- W2099479976 cites W1648445109 @default.
- W2099479976 cites W1746819321 @default.
- W2099479976 cites W1749494163 @default.
- W2099479976 cites W1934021597 @default.
- W2099479976 cites W2005539942 @default.
- W2099479976 cites W2011873184 @default.
- W2099479976 cites W2018173956 @default.
- W2099479976 cites W2040355924 @default.
- W2099479976 cites W2047229728 @default.
- W2099479976 cites W2048511921 @default.
- W2099479976 cites W2066612702 @default.
- W2099479976 cites W2072555316 @default.
- W2099479976 cites W2086883878 @default.
- W2099479976 cites W2097408416 @default.
- W2099479976 cites W2104924585 @default.
- W2099479976 cites W2106108430 @default.
- W2099479976 cites W2108677974 @default.
- W2099479976 cites W2108966602 @default.
- W2099479976 cites W2121823518 @default.
- W2099479976 cites W2126700535 @default.
- W2099479976 cites W2127870457 @default.
- W2099479976 cites W2132152975 @default.
- W2099479976 cites W2132900105 @default.
- W2099479976 cites W2134214176 @default.
- W2099479976 cites W2136981284 @default.
- W2099479976 cites W2138857742 @default.
- W2099479976 cites W2139701068 @default.
- W2099479976 cites W2140966562 @default.
- W2099479976 cites W2141800987 @default.
- W2099479976 cites W2150394223 @default.
- W2099479976 cites W2156140776 @default.
- W2099479976 cites W2157826563 @default.
- W2099479976 cites W2159375150 @default.
- W2099479976 cites W2162191125 @default.
- W2099479976 cites W2166471851 @default.
- W2099479976 cites W2171980229 @default.
- W2099479976 cites W2606321545 @default.
- W2099479976 cites W2949856351 @default.
- W2099479976 cites W2992769304 @default.
- W2099479976 cites W3120740533 @default.
- W2099479976 cites W76947095 @default.
- W2099479976 cites W1583700866 @default.
- W2099479976 hasPublicationYear "2014" @default.
- W2099479976 type Work @default.
- W2099479976 sameAs 2099479976 @default.
- W2099479976 citedByCount "11" @default.
- W2099479976 countsByYear W20994799762014 @default.
- W2099479976 countsByYear W20994799762015 @default.
- W2099479976 countsByYear W20994799762016 @default.
- W2099479976 countsByYear W20994799762017 @default.
- W2099479976 countsByYear W20994799762018 @default.
- W2099479976 countsByYear W20994799762019 @default.
- W2099479976 crossrefType "journal-article" @default.
- W2099479976 hasAuthorship W2099479976A5047269552 @default.
- W2099479976 hasAuthorship W2099479976A5048712208 @default.
- W2099479976 hasAuthorship W2099479976A5075462759 @default.
- W2099479976 hasConcept C105795698 @default.
- W2099479976 hasConcept C107673813 @default.
- W2099479976 hasConcept C111350023 @default.
- W2099479976 hasConcept C11413529 @default.
- W2099479976 hasConcept C126255220 @default.
- W2099479976 hasConcept C153180895 @default.
- W2099479976 hasConcept C154945302 @default.
- W2099479976 hasConcept C177769412 @default.
- W2099479976 hasConcept C22243797 @default.
- W2099479976 hasConcept C33923547 @default.
- W2099479976 hasConcept C41008148 @default.
- W2099479976 hasConcept C49781872 @default.
- W2099479976 hasConcept C57830394 @default.
- W2099479976 hasConcept C9810830 @default.
- W2099479976 hasConceptScore W2099479976C105795698 @default.
- W2099479976 hasConceptScore W2099479976C107673813 @default.
- W2099479976 hasConceptScore W2099479976C111350023 @default.
- W2099479976 hasConceptScore W2099479976C11413529 @default.
- W2099479976 hasConceptScore W2099479976C126255220 @default.
- W2099479976 hasConceptScore W2099479976C153180895 @default.
- W2099479976 hasConceptScore W2099479976C154945302 @default.
- W2099479976 hasConceptScore W2099479976C177769412 @default.
- W2099479976 hasConceptScore W2099479976C22243797 @default.
- W2099479976 hasConceptScore W2099479976C33923547 @default.
- W2099479976 hasConceptScore W2099479976C41008148 @default.
- W2099479976 hasConceptScore W2099479976C49781872 @default.
- W2099479976 hasConceptScore W2099479976C57830394 @default.
- W2099479976 hasConceptScore W2099479976C9810830 @default.
- W2099479976 hasIssue "1" @default.
- W2099479976 hasLocation W20994799761 @default.