Matches in SemOpenAlex for { <https://semopenalex.org/work/W2950115424> ?p ?o ?g. }
Showing items 1 to 88 of
88
with 100 items per page.
- W2950115424 abstract "For many large undirected models that arise in real-world applications, exact maximumlikelihood training is intractable, because it requires computing marginal distributions of the model. Conditional training is even more difficult, because the partition function depends not only on the parameters, but also on the observed input, requiring repeated inference over each training example. An appealing idea for such models is to independently train a local undirected classifier over each clique, afterwards combining the learned weights into a single global model. In this paper, we show that this piecewise method can be justified as minimizing a new family of upper bounds on the log partition function. On three natural-language data sets, piecewise training is more accurate than pseudolikelihood, and often performs comparably to global training using belief propagation." @default.
- W2950115424 created "2019-06-27" @default.
- W2950115424 creator A5008354502 @default.
- W2950115424 creator A5028501178 @default.
- W2950115424 date "2012-07-04" @default.
- W2950115424 modified "2023-09-27" @default.
- W2950115424 title "Piecewise Training for Undirected Models" @default.
- W2950115424 cites W1528056001 @default.
- W2950115424 cites W2129712609 @default.
- W2950115424 cites W2141099517 @default.
- W2950115424 cites W2147880316 @default.
- W2950115424 cites W2152455533 @default.
- W2950115424 cites W2158823144 @default.
- W2950115424 cites W2163915185 @default.
- W2950115424 cites W28766783 @default.
- W2950115424 cites W2962735828 @default.
- W2950115424 cites W78511820 @default.
- W2950115424 hasPublicationYear "2012" @default.
- W2950115424 type Work @default.
- W2950115424 sameAs 2950115424 @default.
- W2950115424 citedByCount "1" @default.
- W2950115424 countsByYear W29501154242018 @default.
- W2950115424 crossrefType "posted-content" @default.
- W2950115424 hasAuthorship W2950115424A5008354502 @default.
- W2950115424 hasAuthorship W2950115424A5028501178 @default.
- W2950115424 hasConcept C11413529 @default.
- W2950115424 hasConcept C114614502 @default.
- W2950115424 hasConcept C119857082 @default.
- W2950115424 hasConcept C126255220 @default.
- W2950115424 hasConcept C132525143 @default.
- W2950115424 hasConcept C134306372 @default.
- W2950115424 hasConcept C154945302 @default.
- W2950115424 hasConcept C164660894 @default.
- W2950115424 hasConcept C2776214188 @default.
- W2950115424 hasConcept C2777035058 @default.
- W2950115424 hasConcept C2777472644 @default.
- W2950115424 hasConcept C3018234147 @default.
- W2950115424 hasConcept C33923547 @default.
- W2950115424 hasConcept C41008148 @default.
- W2950115424 hasConcept C42812 @default.
- W2950115424 hasConcept C51632099 @default.
- W2950115424 hasConcept C80444323 @default.
- W2950115424 hasConcept C95623464 @default.
- W2950115424 hasConceptScore W2950115424C11413529 @default.
- W2950115424 hasConceptScore W2950115424C114614502 @default.
- W2950115424 hasConceptScore W2950115424C119857082 @default.
- W2950115424 hasConceptScore W2950115424C126255220 @default.
- W2950115424 hasConceptScore W2950115424C132525143 @default.
- W2950115424 hasConceptScore W2950115424C134306372 @default.
- W2950115424 hasConceptScore W2950115424C154945302 @default.
- W2950115424 hasConceptScore W2950115424C164660894 @default.
- W2950115424 hasConceptScore W2950115424C2776214188 @default.
- W2950115424 hasConceptScore W2950115424C2777035058 @default.
- W2950115424 hasConceptScore W2950115424C2777472644 @default.
- W2950115424 hasConceptScore W2950115424C3018234147 @default.
- W2950115424 hasConceptScore W2950115424C33923547 @default.
- W2950115424 hasConceptScore W2950115424C41008148 @default.
- W2950115424 hasConceptScore W2950115424C42812 @default.
- W2950115424 hasConceptScore W2950115424C51632099 @default.
- W2950115424 hasConceptScore W2950115424C80444323 @default.
- W2950115424 hasConceptScore W2950115424C95623464 @default.
- W2950115424 hasLocation W29501154241 @default.
- W2950115424 hasOpenAccess W2950115424 @default.
- W2950115424 hasPrimaryLocation W29501154241 @default.
- W2950115424 hasRelatedWork W1608224639 @default.
- W2950115424 hasRelatedWork W167621446 @default.
- W2950115424 hasRelatedWork W1755468095 @default.
- W2950115424 hasRelatedWork W1857926807 @default.
- W2950115424 hasRelatedWork W2001331793 @default.
- W2950115424 hasRelatedWork W2130186484 @default.
- W2950115424 hasRelatedWork W2142165428 @default.
- W2950115424 hasRelatedWork W2153846128 @default.
- W2950115424 hasRelatedWork W2186700413 @default.
- W2950115424 hasRelatedWork W2741878218 @default.
- W2950115424 hasRelatedWork W2947361127 @default.
- W2950115424 hasRelatedWork W2950906236 @default.
- W2950115424 hasRelatedWork W2952793838 @default.
- W2950115424 hasRelatedWork W2964031514 @default.
- W2950115424 hasRelatedWork W3035909655 @default.
- W2950115424 hasRelatedWork W3134450292 @default.
- W2950115424 hasRelatedWork W3163588117 @default.
- W2950115424 hasRelatedWork W3173016912 @default.
- W2950115424 hasRelatedWork W3190275355 @default.
- W2950115424 hasRelatedWork W3198742493 @default.
- W2950115424 isParatext "false" @default.
- W2950115424 isRetracted "false" @default.
- W2950115424 magId "2950115424" @default.
- W2950115424 workType "article" @default.