Can overfitted deep neural networks in adversarial training generalize? - An approximation viewpoint.
Zhongjie ShiFanghui LiuYuan CaoJohan A. K. SuykensPublished in: CoRR (2024)
Keyphrases
- viewpoint
- neural network
- training process
- training algorithm
- feedforward neural networks
- feed forward neural networks
- neural network training
- multi agent
- pattern recognition
- deep architectures
- training set
- multi layer perceptron
- neural network structure
- feed forward
- training phase
- recurrent networks
- training data
- genetic algorithm
- approximation error
- training patterns
- back propagation
- backpropagation algorithm
- hidden layer
- error bounds
- multi layer
- neural nets
- fuzzy logic
- error back propagation
- structured prediction
- multilayer neural network
- fuzzy systems
- multilayer perceptron
- lighting conditions
- closed form
- training examples
- training samples
- supervised learning
- support vector machine
- artificial neural networks
- machine learning