Investigating a spectral deception loss metric for training machine learning-based evasion attacks.
Matthew DelVecchioVanessa ArndorferWilliam C. HeadleyPublished in: WiseML@WiSec (2020)
Keyphrases
- machine learning
- countermeasures
- supervised learning
- training set
- training phase
- incomplete information
- reinforcement learning
- machine learning methods
- online learning
- data mining
- network anomaly detection
- training process
- knowledge acquisition
- computer vision
- pattern recognition
- feature selection
- support vector machine
- data analysis
- computer science
- training algorithm
- test set
- learning tasks
- metric space
- neural network
- covariate shift
- hyperspectral images
- learning algorithm
- metric learning
- decision trees
- machine learning algorithms
- training examples
- text classification
- natural language processing
- active learning
- information extraction