Active Learning Under Malicious Mislabeling and Poisoning Attacks.
Jing LinRyan LuleyKaiqi XiongPublished in: CoRR (2021)
Keyphrases
- active learning
- malicious users
- malicious attacks
- detect malicious
- detecting malicious
- malicious behavior
- insider attacks
- malicious activities
- malicious code
- insider threat
- learning strategies
- machine learning
- countermeasures
- normal traffic
- random sampling
- selective sampling
- malicious traffic
- semi fragile watermarking
- learning algorithm
- semi supervised
- malicious nodes
- experimental design
- website
- learning process
- computer security
- relevance feedback
- supervised learning
- training examples
- identity theft
- watermarking scheme
- end hosts
- labeled data
- traffic analysis
- security threats
- batch mode
- network traffic
- attack detection
- wireless sensor networks
- data sets
- imbalanced data classification
- intrusion detection system
- watermarking algorithm
- training data
- training set
- sensitive data
- active learning strategies
- static analysis
- denial of service attacks
- anomaly detection