Is Approximation Universally Defensive Against Adversarial Attacks in Deep Neural Networks?
Ayesha SiddiqueKhaza Anuarul HoquePublished in: DATE (2022)
Keyphrases
- web services
- neural network
- pattern recognition
- countermeasures
- artificial neural networks
- approximation error
- watermarking scheme
- neural nets
- back propagation
- fuzzy logic
- traffic analysis
- fault diagnosis
- error bounds
- neural network model
- approximation methods
- malicious users
- recurrent neural networks
- deep learning
- security protocols
- watermarking method
- relative error
- security mechanisms
- multi layer perceptron
- training process
- fuzzy systems
- smart card
- closed form
- self organizing maps
- multi agent
- learning algorithm
- genetic algorithm