Interpretable Anomaly Detection with DIFFI: Depth-based Feature Importance for the Isolation Forest.
Mattia CarlettiMatteo TerziGian Antonio SustoPublished in: CoRR (2020)
Keyphrases
- anomaly detection
- feature importance
- random forest
- intrusion detection
- detecting anomalies
- network intrusion detection
- network traffic
- unsupervised anomaly detection
- network anomaly detection
- detect anomalies
- anomalous behavior
- detecting anomalous
- feature selection
- relevance feedback
- one class support vector machines
- semi supervised
- intrusion detection system
- network security
- unsupervised learning
- normal behavior
- network intrusion
- image segmentation
- decision trees
- cumulative sum