Post-Robustifying Deep Anomaly Detection Ensembles by Model Selection.
Benedikt BöingSimon KlüttermannEmmanuel MüllerPublished in: ICDM (2022)
Keyphrases
- model selection
- anomaly detection
- cross validation
- unsupervised learning
- intrusion detection
- anomalous behavior
- hyperparameters
- detecting anomalies
- bayesian learning
- network traffic
- network intrusion detection
- network anomaly detection
- generalization error
- variable selection
- machine learning
- model selection criteria
- decision trees
- information criterion
- selection criterion
- feature selection
- statistical learning
- ensemble learning
- data sets
- ensemble methods
- intrusion detection system
- bayesian information criterion
- one class support vector machines
- pattern recognition
- data mining techniques
- automatic model selection
- detect anomalies
- marginal likelihood
- maximum likelihood
- clustering algorithm
- real world