Removing Batch Normalization Boosts Adversarial Training.
Haotao WangAston ZhangShuai ZhengXingjian ShiMu LiZhangyang WangPublished in: CoRR (2022)
Keyphrases
- batch mode
- training process
- training set
- information retrieval
- training algorithm
- test set
- training samples
- active learning
- feedforward neural networks
- online learning
- supervised learning
- support vector machine
- data sets
- pairwise
- feature extraction
- case study
- decision making
- learning algorithm
- machine learning
- databases