Multi-Attribute Monitoring for Anomaly Detection: a Reinforcement Learning Approach based on Unsupervised Reward.
Mohamed Said FrikhaSonia Mettali GammarAbdelkader LahmadiPublished in: PEMWN (2021)
Keyphrases
- anomaly detection
- multi attribute
- reinforcement learning
- unsupervised learning
- unsupervised anomaly detection
- detect anomalies
- intrusion detection
- utility function
- supervised learning
- detecting anomalies
- state space
- anomalous behavior
- multiple attributes
- utility theory
- network intrusion detection
- network traffic
- semi supervised
- multi attribute decision making
- attribute values
- learning algorithm
- machine learning
- multi dimensional
- multi agent
- intrusion detection system
- markov decision processes
- optimal policy
- auction mechanisms
- similarity search
- network anomaly detection
- negative selection algorithm
- reward function
- maximum likelihood
- abnormal events
- evolutionary algorithm
- pattern recognition
- one class support vector machines