A Theoretical Framework for Robustness of (Deep) Classifiers against Adversarial Samples.
Beilun WangJi GaoYanjun QiPublished in: ICLR (Workshop) (2017)
Keyphrases
- theoretical framework
- training samples
- optimum path forest
- training set
- small sample
- data samples
- training examples
- theoretical foundation
- multi agent
- support vector
- training data
- conceptual framework
- data sets
- supervised classification
- machine learning algorithms
- fundamental principles
- naive bayes
- information systems
- statistical learning theory
- conceptual change
- collaborative knowledge building
- e government
- supervised learning
- feature selection
- feature space
- learning algorithm
- minority class
- machine learning