Defending against adversarial attacks on graph neural networks via similarity property.
Minghong YaoHaizheng YuHong BianPublished in: AI Commun. (2023)
Keyphrases
- neural network
- ddos attacks
- similarity measure
- global consistency
- graph representation
- graph properties
- similarity graph
- graph theory
- inexact graph matching
- pattern recognition
- random walk
- similarity scores
- distance measure
- normalized cut
- graph model
- artificial neural networks
- fuzzy logic
- graph matching
- graph theoretic
- similarity measurement
- similarity metric
- feed forward
- directed graph
- genetic algorithm
- anti monotonic
- back propagation
- data mining
- security threats
- graph partitioning
- graph structure
- semantic similarity
- euclidean distance
- similarity matrix
- watermarking technique
- recurrent neural networks
- similarity function
- topological information
- self organizing maps
- fault diagnosis
- triangular inequality
- distance function