First-Order Adversarial Vulnerability of Neural Networks and Input Dimension.
Carl-Johann Simon-GabrielYann OllivierLéon BottouBernhard SchölkopfDavid Lopez-PazPublished in: ICML (2019)
Keyphrases
- neural network
- pattern recognition
- desired output
- first order logic
- fuzzy logic
- higher order
- artificial neural networks
- neural network model
- real time
- temporal dimension
- fuzzy systems
- feed forward
- fault diagnosis
- genetic algorithm
- artificial intelligence
- risk assessment
- learning rules
- nonlinear functions
- feedforward neural networks
- back propagation