Matches in SemOpenAlex for { <https://semopenalex.org/work/W3008963359> ?p ?o ?g. }
Showing items 1 to 82 of
82
with 100 items per page.
- W3008963359 abstract "Graph Neural Networks (GNNs) have made significant advances on several fundamental inference tasks. As a result, there is a surge of interest in using these models for making potentially important decisions in high-regret applications. However, despite GNNs' impressive performance, it has been observed that carefully crafted perturbations on graph structures (or nodes attributes) lead them to make wrong predictions. Presence of these adversarial examples raises serious security concerns. Most of the existing robust GNN design/training methods are only applicable to white-box settings where model parameters are known and gradient based methods can be used by performing convex relaxation of the discrete graph domain. More importantly, these methods are not efficient and scalable which make them infeasible in time sensitive tasks and massive graph datasets. To overcome these limitations, we propose a general framework which leverages the greedy search algorithms and zeroth-order methods to obtain robust GNNs in a generic and an efficient manner. On several applications, we show that the proposed techniques are significantly less computationally expensive and, in some cases, more robust than the state-of-the-art methods making them suitable to large-scale problems which were out of the reach of traditional robust training methods." @default.
- W3008963359 created "2020-03-06" @default.
- W3008963359 creator A5002976916 @default.
- W3008963359 creator A5011173717 @default.
- W3008963359 creator A5030060072 @default.
- W3008963359 creator A5041470575 @default.
- W3008963359 creator A5043582832 @default.
- W3008963359 creator A5050344371 @default.
- W3008963359 creator A5083232044 @default.
- W3008963359 date "2020-02-25" @default.
- W3008963359 modified "2023-09-27" @default.
- W3008963359 title "Towards an Efficient and General Framework of Robust Training for Graph Neural Networks" @default.
- W3008963359 cites W2153959628 @default.
- W3008963359 cites W2315403234 @default.
- W3008963359 cites W2519887557 @default.
- W3008963359 cites W2798598284 @default.
- W3008963359 cites W2803678876 @default.
- W3008963359 cites W2803831897 @default.
- W3008963359 cites W2887906331 @default.
- W3008963359 cites W2909079677 @default.
- W3008963359 cites W2914953695 @default.
- W3008963359 cites W2934684307 @default.
- W3008963359 cites W2962711740 @default.
- W3008963359 cites W2963243330 @default.
- W3008963359 cites W2964253222 @default.
- W3008963359 cites W2964971928 @default.
- W3008963359 cites W2969992695 @default.
- W3008963359 cites W2970098918 @default.
- W3008963359 cites W2981892732 @default.
- W3008963359 cites W3009751875 @default.
- W3008963359 cites W3101158652 @default.
- W3008963359 cites W3106390645 @default.
- W3008963359 doi "https://doi.org/10.48550/arxiv.2002.10947" @default.
- W3008963359 hasPublicationYear "2020" @default.
- W3008963359 type Work @default.
- W3008963359 sameAs 3008963359 @default.
- W3008963359 citedByCount "0" @default.
- W3008963359 crossrefType "posted-content" @default.
- W3008963359 hasAuthorship W3008963359A5002976916 @default.
- W3008963359 hasAuthorship W3008963359A5011173717 @default.
- W3008963359 hasAuthorship W3008963359A5030060072 @default.
- W3008963359 hasAuthorship W3008963359A5041470575 @default.
- W3008963359 hasAuthorship W3008963359A5043582832 @default.
- W3008963359 hasAuthorship W3008963359A5050344371 @default.
- W3008963359 hasAuthorship W3008963359A5083232044 @default.
- W3008963359 hasBestOaLocation W30089633591 @default.
- W3008963359 hasConcept C119857082 @default.
- W3008963359 hasConcept C132525143 @default.
- W3008963359 hasConcept C154945302 @default.
- W3008963359 hasConcept C2776214188 @default.
- W3008963359 hasConcept C41008148 @default.
- W3008963359 hasConcept C48044578 @default.
- W3008963359 hasConcept C50817715 @default.
- W3008963359 hasConcept C77088390 @default.
- W3008963359 hasConcept C80444323 @default.
- W3008963359 hasConceptScore W3008963359C119857082 @default.
- W3008963359 hasConceptScore W3008963359C132525143 @default.
- W3008963359 hasConceptScore W3008963359C154945302 @default.
- W3008963359 hasConceptScore W3008963359C2776214188 @default.
- W3008963359 hasConceptScore W3008963359C41008148 @default.
- W3008963359 hasConceptScore W3008963359C48044578 @default.
- W3008963359 hasConceptScore W3008963359C50817715 @default.
- W3008963359 hasConceptScore W3008963359C77088390 @default.
- W3008963359 hasConceptScore W3008963359C80444323 @default.
- W3008963359 hasLocation W30089633591 @default.
- W3008963359 hasLocation W30089633592 @default.
- W3008963359 hasOpenAccess W3008963359 @default.
- W3008963359 hasPrimaryLocation W30089633591 @default.
- W3008963359 hasRelatedWork W10699555 @default.
- W3008963359 hasRelatedWork W11849241 @default.
- W3008963359 hasRelatedWork W12904111 @default.
- W3008963359 hasRelatedWork W14632104 @default.
- W3008963359 hasRelatedWork W14988460 @default.
- W3008963359 hasRelatedWork W2177595 @default.
- W3008963359 hasRelatedWork W2444557 @default.
- W3008963359 hasRelatedWork W2956227 @default.
- W3008963359 hasRelatedWork W4529005 @default.
- W3008963359 hasRelatedWork W8792277 @default.
- W3008963359 isParatext "false" @default.
- W3008963359 isRetracted "false" @default.
- W3008963359 magId "3008963359" @default.
- W3008963359 workType "article" @default.