Matches in SemOpenAlex for { <https://semopenalex.org/work/W4285606429> ?p ?o ?g. }
Showing items 1 to 63 of
63
with 100 items per page.
- W4285606429 abstract "Positive-unlabeled (PU) learning deals with the circumstances where only a portion of positive instances are labeled, while the rest and all negative instances are unlabeled, and due to this confusion, the class prior can not be directly available. Existing PU learning methods usually estimate the class prior by training a nontraditional probabilistic classifier, which is prone to give an overestimation. Moreover, these methods learn the decision boundary by optimizing the minimum margin, which is not suitable in PU learning due to its sensitivity to label noise. In this paper, we enhance PU learning methods from the above two aspects. More specifically, we first explicitly learn a transformation from unlabeled data to positive data by entropy regularized optimal transport to achieve a much more precise estimation for class prior. Then we switch to optimizing the margin distribution, rather than the minimum margin, to obtain a label noise insensitive classifier. Extensive empirical studies on both synthetic and real-world data sets demonstrate the superiority of our proposed method." @default.
- W4285606429 created "2022-07-16" @default.
- W4285606429 creator A5007250349 @default.
- W4285606429 creator A5033785543 @default.
- W4285606429 creator A5065741507 @default.
- W4285606429 creator A5077423550 @default.
- W4285606429 date "2022-07-01" @default.
- W4285606429 modified "2023-09-29" @default.
- W4285606429 title "Posistive-Unlabeled Learning via Optimal Transport and Margin Distribution" @default.
- W4285606429 doi "https://doi.org/10.24963/ijcai.2022/390" @default.
- W4285606429 hasPublicationYear "2022" @default.
- W4285606429 type Work @default.
- W4285606429 citedByCount "0" @default.
- W4285606429 crossrefType "proceedings-article" @default.
- W4285606429 hasAuthorship W4285606429A5007250349 @default.
- W4285606429 hasAuthorship W4285606429A5033785543 @default.
- W4285606429 hasAuthorship W4285606429A5065741507 @default.
- W4285606429 hasAuthorship W4285606429A5077423550 @default.
- W4285606429 hasBestOaLocation W42856064291 @default.
- W4285606429 hasConcept C106301342 @default.
- W4285606429 hasConcept C11171543 @default.
- W4285606429 hasConcept C119857082 @default.
- W4285606429 hasConcept C121332964 @default.
- W4285606429 hasConcept C153180895 @default.
- W4285606429 hasConcept C154945302 @default.
- W4285606429 hasConcept C15744967 @default.
- W4285606429 hasConcept C2781140086 @default.
- W4285606429 hasConcept C41008148 @default.
- W4285606429 hasConcept C42023084 @default.
- W4285606429 hasConcept C49937458 @default.
- W4285606429 hasConcept C62520636 @default.
- W4285606429 hasConcept C774472 @default.
- W4285606429 hasConcept C95623464 @default.
- W4285606429 hasConceptScore W4285606429C106301342 @default.
- W4285606429 hasConceptScore W4285606429C11171543 @default.
- W4285606429 hasConceptScore W4285606429C119857082 @default.
- W4285606429 hasConceptScore W4285606429C121332964 @default.
- W4285606429 hasConceptScore W4285606429C153180895 @default.
- W4285606429 hasConceptScore W4285606429C154945302 @default.
- W4285606429 hasConceptScore W4285606429C15744967 @default.
- W4285606429 hasConceptScore W4285606429C2781140086 @default.
- W4285606429 hasConceptScore W4285606429C41008148 @default.
- W4285606429 hasConceptScore W4285606429C42023084 @default.
- W4285606429 hasConceptScore W4285606429C49937458 @default.
- W4285606429 hasConceptScore W4285606429C62520636 @default.
- W4285606429 hasConceptScore W4285606429C774472 @default.
- W4285606429 hasConceptScore W4285606429C95623464 @default.
- W4285606429 hasLocation W42856064291 @default.
- W4285606429 hasOpenAccess W4285606429 @default.
- W4285606429 hasPrimaryLocation W42856064291 @default.
- W4285606429 hasRelatedWork W2001652754 @default.
- W4285606429 hasRelatedWork W2379065761 @default.
- W4285606429 hasRelatedWork W2549006548 @default.
- W4285606429 hasRelatedWork W2784352036 @default.
- W4285606429 hasRelatedWork W2807311372 @default.
- W4285606429 hasRelatedWork W2961085424 @default.
- W4285606429 hasRelatedWork W2972035100 @default.
- W4285606429 hasRelatedWork W3043252291 @default.
- W4285606429 hasRelatedWork W4214932115 @default.
- W4285606429 hasRelatedWork W3158004940 @default.
- W4285606429 isParatext "false" @default.
- W4285606429 isRetracted "false" @default.
- W4285606429 workType "article" @default.