Matches in SemOpenAlex for { <https://semopenalex.org/work/W4221149697> ?p ?o ?g. }
Showing items 1 to 71 of
71
with 100 items per page.
- W4221149697 abstract "Graph neural networks (GNNs) have been increasingly deployed in various applications that involve learning on non-Euclidean data. However, recent studies show that GNNs are vulnerable to graph adversarial attacks. Although there are several defense methods to improve GNN robustness by eliminating adversarial components, they may also impair the underlying clean graph structure that contributes to GNN training. In addition, few of those defense models can scale to large graphs due to their high computational complexity and memory usage. In this paper, we propose GARNET, a scalable spectral method to boost the adversarial robustness of GNN models. GARNET first leverages weighted spectral embedding to construct a base graph, which is not only resistant to adversarial attacks but also contains critical (clean) graph structure for GNN training. Next, GARNET further refines the base graph by pruning additional uncritical edges based on probabilistic graphical model. GARNET has been evaluated on various datasets, including a large graph with millions of nodes. Our extensive experiment results show that GARNET achieves adversarial accuracy improvement and runtime speedup over state-of-the-art GNN (defense) models by up to 13.27% and 14.7x, respectively." @default.
- W4221149697 created "2022-04-03" @default.
- W4221149697 creator A5020102159 @default.
- W4221149697 creator A5037210004 @default.
- W4221149697 creator A5074490226 @default.
- W4221149697 creator A5079567542 @default.
- W4221149697 date "2022-01-30" @default.
- W4221149697 modified "2023-09-23" @default.
- W4221149697 title "GARNET: Reduced-Rank Topology Learning for Robust and Scalable Graph Neural Networks" @default.
- W4221149697 doi "https://doi.org/10.48550/arxiv.2201.12741" @default.
- W4221149697 hasPublicationYear "2022" @default.
- W4221149697 type Work @default.
- W4221149697 citedByCount "0" @default.
- W4221149697 crossrefType "posted-content" @default.
- W4221149697 hasAuthorship W4221149697A5020102159 @default.
- W4221149697 hasAuthorship W4221149697A5037210004 @default.
- W4221149697 hasAuthorship W4221149697A5074490226 @default.
- W4221149697 hasAuthorship W4221149697A5079567542 @default.
- W4221149697 hasBestOaLocation W42211496971 @default.
- W4221149697 hasConcept C104317684 @default.
- W4221149697 hasConcept C119857082 @default.
- W4221149697 hasConcept C132525143 @default.
- W4221149697 hasConcept C154945302 @default.
- W4221149697 hasConcept C157406716 @default.
- W4221149697 hasConcept C173608175 @default.
- W4221149697 hasConcept C185592680 @default.
- W4221149697 hasConcept C203776342 @default.
- W4221149697 hasConcept C22149727 @default.
- W4221149697 hasConcept C37736160 @default.
- W4221149697 hasConcept C41008148 @default.
- W4221149697 hasConcept C41608201 @default.
- W4221149697 hasConcept C48044578 @default.
- W4221149697 hasConcept C55493867 @default.
- W4221149697 hasConcept C63479239 @default.
- W4221149697 hasConcept C68339613 @default.
- W4221149697 hasConcept C77088390 @default.
- W4221149697 hasConcept C80444323 @default.
- W4221149697 hasConceptScore W4221149697C104317684 @default.
- W4221149697 hasConceptScore W4221149697C119857082 @default.
- W4221149697 hasConceptScore W4221149697C132525143 @default.
- W4221149697 hasConceptScore W4221149697C154945302 @default.
- W4221149697 hasConceptScore W4221149697C157406716 @default.
- W4221149697 hasConceptScore W4221149697C173608175 @default.
- W4221149697 hasConceptScore W4221149697C185592680 @default.
- W4221149697 hasConceptScore W4221149697C203776342 @default.
- W4221149697 hasConceptScore W4221149697C22149727 @default.
- W4221149697 hasConceptScore W4221149697C37736160 @default.
- W4221149697 hasConceptScore W4221149697C41008148 @default.
- W4221149697 hasConceptScore W4221149697C41608201 @default.
- W4221149697 hasConceptScore W4221149697C48044578 @default.
- W4221149697 hasConceptScore W4221149697C55493867 @default.
- W4221149697 hasConceptScore W4221149697C63479239 @default.
- W4221149697 hasConceptScore W4221149697C68339613 @default.
- W4221149697 hasConceptScore W4221149697C77088390 @default.
- W4221149697 hasConceptScore W4221149697C80444323 @default.
- W4221149697 hasLocation W42211496971 @default.
- W4221149697 hasOpenAccess W4221149697 @default.
- W4221149697 hasPrimaryLocation W42211496971 @default.
- W4221149697 hasRelatedWork W157504995 @default.
- W4221149697 hasRelatedWork W1976363619 @default.
- W4221149697 hasRelatedWork W2953770707 @default.
- W4221149697 hasRelatedWork W3044158376 @default.
- W4221149697 hasRelatedWork W3190313451 @default.
- W4221149697 hasRelatedWork W3208308319 @default.
- W4221149697 hasRelatedWork W333278977 @default.
- W4221149697 hasRelatedWork W4297895859 @default.
- W4221149697 hasRelatedWork W4302283439 @default.
- W4221149697 hasRelatedWork W4306909306 @default.
- W4221149697 isParatext "false" @default.
- W4221149697 isRetracted "false" @default.
- W4221149697 workType "article" @default.