Sign in

Bandits for Structure Perturbation-based Black-box Attacks to Graph Neural Networks with Theoretical Guarantees.

Binghui WangYouqi LiPan Zhou
Published in: CVPR (2022)
Keyphrases
  • black box
  • theoretical guarantees
  • neural network
  • rule extraction
  • graph structure
  • hybrid systems
  • black boxes
  • white box
  • integration testing
  • worst case
  • open source
  • test cases
  • white box testing