Unsupervised feature selection using chronological fitting with Shapley Additive explanation (SHAP) for industrial time-series anomaly detection.
Qixuan LiYangjian JiMingrui ZhuXiaoyang ZhuLinjin SunPublished in: Appl. Soft Comput. (2024)
Keyphrases
- anomaly detection
- unsupervised feature selection
- detect anomalies
- intrusion detection
- detecting anomalies
- feature selection
- anomalous behavior
- network traffic
- data clustering
- network intrusion detection
- intrusion detection system
- network anomaly detection
- dimensionality reduction
- one class support vector machines
- unsupervised learning
- negative selection algorithm
- least squares
- active learning
- feature vectors
- object recognition