Demystifying Uneven Vulnerability of Link Stealing Attacks against Graph Neural Networks.
He ZhangBang WuShuo WangXiangwen YangMinhui XueShirui PanXingliang YuanPublished in: ICML (2023)
Keyphrases
- neural network
- buffer overflow
- graph structure
- security risks
- graph model
- graph representation
- pattern recognition
- fuzzy logic
- random walk
- neural network model
- graph theory
- web graph
- graph theoretic
- denial of service
- dos attacks
- connected components
- bipartite graph
- fault diagnosis
- security vulnerabilities
- graph mining
- multi layer
- malicious attacks
- countermeasures
- genetic algorithm
- artificial neural networks
- neural nets
- back propagation
- recurrent neural networks
- graph matching
- feed forward
- directed graph
- structured data
- security threats
- malicious users
- attack detection
- strongly connected
- link analysis
- link structure
- graph partitioning
- weighted graph
- undirected graph