GradDiv: Adversarial Robustness of Randomized Neural Networks via Gradient Diversity Regularization.
Sungyoon LeeHoki KimJaewook LeePublished in: CoRR (2021)
Keyphrases
- neural network
- pattern recognition
- artificial neural networks
- fuzzy logic
- back propagation
- multi layer
- prior information
- neural network model
- gradient direction
- multi agent
- multilayer perceptron
- neural nets
- fault diagnosis
- activation function
- training process
- radial basis function
- multiscale
- decision trees
- genetic algorithm