Matches in SemOpenAlex for { <https://semopenalex.org/work/W4293869698> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4293869698 abstract "The standard margin-based structured prediction commonly uses a maximum loss over all possible structured outputs. The large-margin formulation including latent variables not only results in a non-convex formulation but also increases the search space by a factor of the size of the latent space. Recent work has proposed the use of the maximum loss over random structured outputs sampled independently from some proposal distribution, with theoretical guarantees. We extend this work by including latent variables. We study a new family of loss functions under Gaussian perturbations and analyze the effect of the latent space on the generalization bounds. We show that the non-convexity of learning with latent variables originates naturally, as it relates to a tight upper bound of the Gibbs decoder distortion with respect to the latent space. Finally, we provide a formulation using random samples that produces a tighter upper bound of the Gibbs decoder distortion up to a statistical accuracy, which enables a faster evaluation of the objective function. We illustrate the method with synthetic experiments and a computer vision application." @default.
- W4293869698 created "2022-09-01" @default.
- W4293869698 creator A5018564606 @default.
- W4293869698 creator A5038965118 @default.
- W4293869698 date "2018-05-23" @default.
- W4293869698 modified "2023-10-16" @default.
- W4293869698 title "Learning latent variable structured prediction models with Gaussian perturbations" @default.
- W4293869698 doi "https://doi.org/10.48550/arxiv.1805.09213" @default.
- W4293869698 hasPublicationYear "2018" @default.
- W4293869698 type Work @default.
- W4293869698 citedByCount "0" @default.
- W4293869698 crossrefType "posted-content" @default.
- W4293869698 hasAuthorship W4293869698A5018564606 @default.
- W4293869698 hasAuthorship W4293869698A5038965118 @default.
- W4293869698 hasBestOaLocation W42938696981 @default.
- W4293869698 hasConcept C105795698 @default.
- W4293869698 hasConcept C106159729 @default.
- W4293869698 hasConcept C11413529 @default.
- W4293869698 hasConcept C119857082 @default.
- W4293869698 hasConcept C121332964 @default.
- W4293869698 hasConcept C122123141 @default.
- W4293869698 hasConcept C126255220 @default.
- W4293869698 hasConcept C126780896 @default.
- W4293869698 hasConcept C134306372 @default.
- W4293869698 hasConcept C162324750 @default.
- W4293869698 hasConcept C163716315 @default.
- W4293869698 hasConcept C177148314 @default.
- W4293869698 hasConcept C194257627 @default.
- W4293869698 hasConcept C2776257435 @default.
- W4293869698 hasConcept C28826006 @default.
- W4293869698 hasConcept C31258907 @default.
- W4293869698 hasConcept C33923547 @default.
- W4293869698 hasConcept C41008148 @default.
- W4293869698 hasConcept C51167844 @default.
- W4293869698 hasConcept C62520636 @default.
- W4293869698 hasConcept C65965080 @default.
- W4293869698 hasConcept C72134830 @default.
- W4293869698 hasConcept C774472 @default.
- W4293869698 hasConcept C77553402 @default.
- W4293869698 hasConceptScore W4293869698C105795698 @default.
- W4293869698 hasConceptScore W4293869698C106159729 @default.
- W4293869698 hasConceptScore W4293869698C11413529 @default.
- W4293869698 hasConceptScore W4293869698C119857082 @default.
- W4293869698 hasConceptScore W4293869698C121332964 @default.
- W4293869698 hasConceptScore W4293869698C122123141 @default.
- W4293869698 hasConceptScore W4293869698C126255220 @default.
- W4293869698 hasConceptScore W4293869698C126780896 @default.
- W4293869698 hasConceptScore W4293869698C134306372 @default.
- W4293869698 hasConceptScore W4293869698C162324750 @default.
- W4293869698 hasConceptScore W4293869698C163716315 @default.
- W4293869698 hasConceptScore W4293869698C177148314 @default.
- W4293869698 hasConceptScore W4293869698C194257627 @default.
- W4293869698 hasConceptScore W4293869698C2776257435 @default.
- W4293869698 hasConceptScore W4293869698C28826006 @default.
- W4293869698 hasConceptScore W4293869698C31258907 @default.
- W4293869698 hasConceptScore W4293869698C33923547 @default.
- W4293869698 hasConceptScore W4293869698C41008148 @default.
- W4293869698 hasConceptScore W4293869698C51167844 @default.
- W4293869698 hasConceptScore W4293869698C62520636 @default.
- W4293869698 hasConceptScore W4293869698C65965080 @default.
- W4293869698 hasConceptScore W4293869698C72134830 @default.
- W4293869698 hasConceptScore W4293869698C774472 @default.
- W4293869698 hasConceptScore W4293869698C77553402 @default.
- W4293869698 hasLocation W42938696981 @default.
- W4293869698 hasOpenAccess W4293869698 @default.
- W4293869698 hasPrimaryLocation W42938696981 @default.
- W4293869698 hasRelatedWork W1994912998 @default.
- W4293869698 hasRelatedWork W2074390224 @default.
- W4293869698 hasRelatedWork W2092244978 @default.
- W4293869698 hasRelatedWork W2362200860 @default.
- W4293869698 hasRelatedWork W2616125534 @default.
- W4293869698 hasRelatedWork W2770703741 @default.
- W4293869698 hasRelatedWork W2783438107 @default.
- W4293869698 hasRelatedWork W2804526847 @default.
- W4293869698 hasRelatedWork W3099524215 @default.
- W4293869698 hasRelatedWork W4293869698 @default.
- W4293869698 isParatext "false" @default.
- W4293869698 isRetracted "false" @default.
- W4293869698 workType "article" @default.