Matches in SemOpenAlex for { <https://semopenalex.org/work/W3156989090> ?p ?o ?g. }
- W3156989090 abstract "Graph Neural Networks (GNNs) achieve significant performance for various learning tasks on geometric data due to the incorporation of graph structure into the learning of node representations, which renders their comprehension challenging. In this paper, we first propose a unified framework satisfied by most existing GNN explainers. Then, we introduce GraphSVX, a post hoc local model-agnostic explanation method specifically designed for GNNs. GraphSVX is a decomposition technique that captures the fair contribution of each feature and node towards the explained prediction by constructing a surrogate model on a perturbed dataset. It extends to graphs and ultimately provides as explanation the Shapley Values from game theory. Experiments on real-world and synthetic datasets demonstrate that GraphSVX achieves state-of-the-art performance compared to baseline models while presenting core theoretical and human-centric properties." @default.
- W3156989090 created "2021-04-26" @default.
- W3156989090 creator A5083981921 @default.
- W3156989090 creator A5087251228 @default.
- W3156989090 date "2021-04-18" @default.
- W3156989090 modified "2023-09-27" @default.
- W3156989090 title "GraphSVX: Shapley Value Explanations for Graph Neural Networks" @default.
- W3156989090 cites W1492230849 @default.
- W3156989090 cites W1556803204 @default.
- W3156989090 cites W1562353621 @default.
- W3156989090 cites W1971916086 @default.
- W3156989090 cites W2014181466 @default.
- W3156989090 cites W2107569009 @default.
- W3156989090 cites W2110953678 @default.
- W3156989090 cites W2125847307 @default.
- W3156989090 cites W2129888542 @default.
- W3156989090 cites W2150480892 @default.
- W3156989090 cites W2152052288 @default.
- W3156989090 cites W2282821441 @default.
- W3156989090 cites W2467510144 @default.
- W3156989090 cites W2510508396 @default.
- W3156989090 cites W2519887557 @default.
- W3156989090 cites W2573660794 @default.
- W3156989090 cites W2605409611 @default.
- W3156989090 cites W2774522520 @default.
- W3156989090 cites W2776207810 @default.
- W3156989090 cites W2805516822 @default.
- W3156989090 cites W2807124908 @default.
- W3156989090 cites W2905224888 @default.
- W3156989090 cites W2947873642 @default.
- W3156989090 cites W2962767366 @default.
- W3156989090 cites W2962851944 @default.
- W3156989090 cites W2962858109 @default.
- W3156989090 cites W2962862931 @default.
- W3156989090 cites W2963095307 @default.
- W3156989090 cites W2963715038 @default.
- W3156989090 cites W2963757395 @default.
- W3156989090 cites W2963858333 @default.
- W3156989090 cites W2964113829 @default.
- W3156989090 cites W2964321699 @default.
- W3156989090 cites W2964811671 @default.
- W3156989090 cites W2979481854 @default.
- W3156989090 cites W3000120900 @default.
- W3156989090 cites W3011667710 @default.
- W3156989090 cites W3101026663 @default.
- W3156989090 cites W3103717137 @default.
- W3156989090 cites W3105503635 @default.
- W3156989090 cites W3116637551 @default.
- W3156989090 cites W3122172846 @default.
- W3156989090 cites W3128452405 @default.
- W3156989090 cites W2970119293 @default.
- W3156989090 hasPublicationYear "2021" @default.
- W3156989090 type Work @default.
- W3156989090 sameAs 3156989090 @default.
- W3156989090 citedByCount "1" @default.
- W3156989090 countsByYear W31569890902021 @default.
- W3156989090 crossrefType "posted-content" @default.
- W3156989090 hasAuthorship W3156989090A5083981921 @default.
- W3156989090 hasAuthorship W3156989090A5087251228 @default.
- W3156989090 hasConcept C119857082 @default.
- W3156989090 hasConcept C127413603 @default.
- W3156989090 hasConcept C132525143 @default.
- W3156989090 hasConcept C144237770 @default.
- W3156989090 hasConcept C154945302 @default.
- W3156989090 hasConcept C177142836 @default.
- W3156989090 hasConcept C199022921 @default.
- W3156989090 hasConcept C2164484 @default.
- W3156989090 hasConcept C2984842247 @default.
- W3156989090 hasConcept C33923547 @default.
- W3156989090 hasConcept C41008148 @default.
- W3156989090 hasConcept C50644808 @default.
- W3156989090 hasConcept C62611344 @default.
- W3156989090 hasConcept C66938386 @default.
- W3156989090 hasConcept C76155785 @default.
- W3156989090 hasConcept C80444323 @default.
- W3156989090 hasConceptScore W3156989090C119857082 @default.
- W3156989090 hasConceptScore W3156989090C127413603 @default.
- W3156989090 hasConceptScore W3156989090C132525143 @default.
- W3156989090 hasConceptScore W3156989090C144237770 @default.
- W3156989090 hasConceptScore W3156989090C154945302 @default.
- W3156989090 hasConceptScore W3156989090C177142836 @default.
- W3156989090 hasConceptScore W3156989090C199022921 @default.
- W3156989090 hasConceptScore W3156989090C2164484 @default.
- W3156989090 hasConceptScore W3156989090C2984842247 @default.
- W3156989090 hasConceptScore W3156989090C33923547 @default.
- W3156989090 hasConceptScore W3156989090C41008148 @default.
- W3156989090 hasConceptScore W3156989090C50644808 @default.
- W3156989090 hasConceptScore W3156989090C62611344 @default.
- W3156989090 hasConceptScore W3156989090C66938386 @default.
- W3156989090 hasConceptScore W3156989090C76155785 @default.
- W3156989090 hasConceptScore W3156989090C80444323 @default.
- W3156989090 hasLocation W31569890901 @default.
- W3156989090 hasOpenAccess W3156989090 @default.
- W3156989090 hasPrimaryLocation W31569890901 @default.
- W3156989090 hasRelatedWork W1723636858 @default.
- W3156989090 hasRelatedWork W2768532803 @default.
- W3156989090 hasRelatedWork W2963067999 @default.
- W3156989090 hasRelatedWork W2991260259 @default.
- W3156989090 hasRelatedWork W3016124664 @default.
- W3156989090 hasRelatedWork W3025414823 @default.