Adversarial attack detection on node classification by autoencoder-based analysis of hidden layers in graph convolutional networks.
Kenta ShimadaTomonobu OzakiPublished in: CANDAR (Workshops) (2021)
Keyphrases
- restricted boltzmann machine
- feedforward neural networks
- classification accuracy
- attack detection
- support vector
- hidden layer
- conditional random fields
- support vector machine svm
- supervised learning
- neural nets
- fault diagnosis
- text categorization
- genetic algorithm
- feature space
- data analysis
- pattern recognition
- feature extraction
- decision making
- feature selection