Few-shot anomaly detection using positive unlabeled learning with cycle consistency and co-occurrence features.
Sion AnJaehong KimSoopil KimPhilip ChikontweJiwook JungHyejeong JeonSang Hyun ParkPublished in: Expert Syst. Appl. (2024)
Keyphrases
- co occurrence
- anomaly detection
- unsupervised learning
- intrusion detection
- positive and unlabeled examples
- connectionist systems
- wordnet
- active learning
- learning process
- semantic relations
- anomalous behavior
- detecting anomalies
- named entities
- learning algorithm
- supervised learning
- network traffic
- network intrusion detection
- feature extraction
- feature set
- information retrieval
- visual features
- feature vectors
- training data