Explanation-Guided Backdoor Poisoning Attacks Against Malware Classifiers.
Giorgio SeveriJim MeyerScott E. CoullAlina OpreaPublished in: USENIX Security Symposium (2021)
Keyphrases
- detect malicious
- malicious code
- training data
- detecting malicious
- decision trees
- attack scenarios
- static analysis
- support vector
- multiple classifiers
- reverse engineering
- linear classifiers
- malware detection
- training set
- test set
- countermeasures
- watermarking scheme
- classifier ensemble
- supervised classification
- classification models
- training samples
- feature selection
- data sets
- machine learning methods
- computer security
- class labels
- naive bayes
- anti virus
- attack detection
- source code
- digital images