FLAGS: A methodology for adaptive anomaly detection and root cause analysis on sensor data streams by fusing expert knowledge with machine learning.
Bram SteenwinckelDieter De PaepeSander Vanden HauttePieter HeyvaertMohamed BentefritPieter MoensAnastasia DimouBruno Van Den BosscheFilip De TurckSofie Van HoeckeFemke OngenaePublished in: Future Gener. Comput. Syst. (2021)
Keyphrases
- anomaly detection
- expert knowledge
- root cause analysis
- machine learning
- expert systems
- network anomaly detection
- intrusion detection
- domain knowledge
- sensor data streams
- detecting anomalies
- bayesian networks
- anomalous behavior
- fuzzy logic
- knowledge base
- network traffic
- domain experts
- network intrusion detection
- one class support vector machines
- prior knowledge
- learning algorithm
- unsupervised learning
- decision trees
- detect anomalies
- knowledge acquisition
- intrusion detection system
- pattern recognition
- knowledge representation
- decision support
- data analysis
- text mining
- machine learning algorithms
- negative selection algorithm
- machine learning methods
- neural network
- computer vision
- root cause
- abnormal events
- data mining
- artificial intelligence
- active learning
- feature selection
- reinforcement learning
- sensor readings
- hidden markov models
- supervised learning
- decision support system
- model selection