Matches in SemOpenAlex for { <https://semopenalex.org/work/W3185981102> ?p ?o ?g. }
Showing items 1 to 89 of
89
with 100 items per page.
- W3185981102 abstract "Graph Neural Networks (GNNs) exhibit excellent performance when graphs have strong homophily property, i.e. connected nodes have the same labels. However, they perform poorly on heterophilic graphs. Several approaches address the issue of heterophily by proposing models that adapt the graph by optimizing task-specific loss function using labelled data. These adaptations are made either via attention or by attenuating or enhancing various low-frequency/high-frequency signals, as needed for the task at hand. More recent approaches adapt the eigenvalues of the graph. One important interpretation of this adaptation is that these models select/weigh the eigenvectors of the graph. Based on this interpretation, we present an eigendecomposition based approach and propose EigenNetwork models that improve the performance of GNNs on heterophilic graphs. Performance improvement is achieved by learning flexible graph adaptation functions that modulate the eigenvalues of the graph. Regularization of these functions via parameter sharing helps to improve the performance even more. Our approach achieves up to 11% improvement in performance over the state-of-the-art methods on heterophilic graphs." @default.
- W3185981102 created "2021-08-02" @default.
- W3185981102 creator A5049226714 @default.
- W3185981102 creator A5057449678 @default.
- W3185981102 creator A5090165776 @default.
- W3185981102 creator A5090842155 @default.
- W3185981102 date "2021-07-28" @default.
- W3185981102 modified "2023-09-27" @default.
- W3185981102 title "Effective Eigendecomposition based Graph Adaptation for Heterophilic Networks." @default.
- W3185981102 cites W1585529040 @default.
- W3185981102 cites W2112796928 @default.
- W3185981102 cites W2130354913 @default.
- W3185981102 cites W2624431344 @default.
- W3185981102 cites W2626778328 @default.
- W3185981102 cites W2784814091 @default.
- W3185981102 cites W2916106175 @default.
- W3185981102 cites W2964121744 @default.
- W3185981102 cites W2964311892 @default.
- W3185981102 cites W2964321699 @default.
- W3185981102 cites W3005644236 @default.
- W3185981102 cites W3120233338 @default.
- W3185981102 cites W3122063025 @default.
- W3185981102 cites W3160872503 @default.
- W3185981102 cites W3173240144 @default.
- W3185981102 cites W3176085777 @default.
- W3185981102 hasPublicationYear "2021" @default.
- W3185981102 type Work @default.
- W3185981102 sameAs 3185981102 @default.
- W3185981102 citedByCount "0" @default.
- W3185981102 crossrefType "posted-content" @default.
- W3185981102 hasAuthorship W3185981102A5049226714 @default.
- W3185981102 hasAuthorship W3185981102A5057449678 @default.
- W3185981102 hasAuthorship W3185981102A5090165776 @default.
- W3185981102 hasAuthorship W3185981102A5090842155 @default.
- W3185981102 hasConcept C114614502 @default.
- W3185981102 hasConcept C119857082 @default.
- W3185981102 hasConcept C120665830 @default.
- W3185981102 hasConcept C121332964 @default.
- W3185981102 hasConcept C132525143 @default.
- W3185981102 hasConcept C139807058 @default.
- W3185981102 hasConcept C154945302 @default.
- W3185981102 hasConcept C158693339 @default.
- W3185981102 hasConcept C169756996 @default.
- W3185981102 hasConcept C2779812341 @default.
- W3185981102 hasConcept C33923547 @default.
- W3185981102 hasConcept C41008148 @default.
- W3185981102 hasConcept C62520636 @default.
- W3185981102 hasConcept C80444323 @default.
- W3185981102 hasConceptScore W3185981102C114614502 @default.
- W3185981102 hasConceptScore W3185981102C119857082 @default.
- W3185981102 hasConceptScore W3185981102C120665830 @default.
- W3185981102 hasConceptScore W3185981102C121332964 @default.
- W3185981102 hasConceptScore W3185981102C132525143 @default.
- W3185981102 hasConceptScore W3185981102C139807058 @default.
- W3185981102 hasConceptScore W3185981102C154945302 @default.
- W3185981102 hasConceptScore W3185981102C158693339 @default.
- W3185981102 hasConceptScore W3185981102C169756996 @default.
- W3185981102 hasConceptScore W3185981102C2779812341 @default.
- W3185981102 hasConceptScore W3185981102C33923547 @default.
- W3185981102 hasConceptScore W3185981102C41008148 @default.
- W3185981102 hasConceptScore W3185981102C62520636 @default.
- W3185981102 hasConceptScore W3185981102C80444323 @default.
- W3185981102 hasLocation W31859811021 @default.
- W3185981102 hasOpenAccess W3185981102 @default.
- W3185981102 hasPrimaryLocation W31859811021 @default.
- W3185981102 hasRelatedWork W2899379687 @default.
- W3185981102 hasRelatedWork W2899886700 @default.
- W3185981102 hasRelatedWork W2945848398 @default.
- W3185981102 hasRelatedWork W3092774975 @default.
- W3185981102 hasRelatedWork W3101198795 @default.
- W3185981102 hasRelatedWork W3102154670 @default.
- W3185981102 hasRelatedWork W3116239416 @default.
- W3185981102 hasRelatedWork W3119224945 @default.
- W3185981102 hasRelatedWork W3120233338 @default.
- W3185981102 hasRelatedWork W3153880451 @default.
- W3185981102 hasRelatedWork W3156642753 @default.
- W3185981102 hasRelatedWork W3164968330 @default.
- W3185981102 hasRelatedWork W3166679531 @default.
- W3185981102 hasRelatedWork W3177075738 @default.
- W3185981102 hasRelatedWork W3189231572 @default.
- W3185981102 hasRelatedWork W3204146977 @default.
- W3185981102 hasRelatedWork W3204410986 @default.
- W3185981102 hasRelatedWork W3209248212 @default.
- W3185981102 hasRelatedWork W3210288927 @default.
- W3185981102 hasRelatedWork W3210699097 @default.
- W3185981102 isParatext "false" @default.
- W3185981102 isRetracted "false" @default.
- W3185981102 magId "3185981102" @default.
- W3185981102 workType "article" @default.