On-line anomaly detection with advanced independent component analysis of multi-variate residual signals from causal relation networks.
Edwin LughoferAlexandru-Ciprian ZavoianuRobert PollakMahardhika PratamaPauline Meyer-HeyeHelmut ZörrerChristian EitzingerThomas RadauerPublished in: Inf. Sci. (2020)
Keyphrases
- independent component analysis
- anomaly detection
- multi variate
- independent components
- signal processing
- intrusion detection
- blind source separation
- principal component analysis
- factor analysis
- detecting anomalies
- anomalous behavior
- blind signal separation
- one class support vector machines
- source separation
- network traffic
- blind separation
- gaussian distribution
- source signals
- intrusion detection system
- principle component analysis
- single channel
- unsupervised learning
- eeg data
- high order statistics
- causal relationships
- detect anomalies
- high dimensional
- feature extraction