A General Backdoor Attack to Graph Neural Networks Based on Explanation Method.
Luyao ChenNa YanBoyang ZhangZhaoyang WangYu WenYanfei HuPublished in: TrustCom (2022)
Keyphrases
- neural network
- detection method
- high accuracy
- segmentation method
- pairwise
- significant improvement
- special case
- dynamic programming
- similarity measure
- directed graph
- computational cost
- clustering method
- graph representation
- neural nets
- probabilistic model
- artificial neural networks
- feature space
- pattern recognition
- image segmentation
- social networks