Heterogeneous Graph Neural Networks for Malicious Account Detection.
Ziqi LiuChaochao ChenXinxing YangJun ZhouXiaolong LiLe SongPublished in: CoRR (2020)
Keyphrases
- neural network
- detection method
- false alarms
- genetic algorithm
- automatic detection
- pattern recognition
- detection accuracy
- random walk
- false positives
- object detection
- detection algorithm
- artificial neural networks
- graph structure
- graph model
- multi layer
- graph representation
- multilayer perceptron
- neural nets
- connected components
- back propagation
- fuzzy logic
- recurrent neural networks
- graph theory
- detection rate
- directed graph
- link analysis
- bayesian networks
- graph theoretic
- graph data
- random graphs
- malicious attacks
- malicious behavior
- normal traffic