Semi-supervised Learning from Active Noisy Soft Labels for Anomaly Detection.
Timo MartensLorenzo PeriniJesse DavisPublished in: ECML/PKDD (1) (2023)
Keyphrases
- anomaly detection
- semi supervised learning
- label propagation
- unsupervised learning
- labeled examples
- semi supervised
- multi instance learning
- unlabeled data
- training data
- labeled data
- intrusion detection
- partially labeled
- unlabeled samples
- supervised learning
- detecting anomalies
- intrusion detection system
- machine learning
- semi supervised classification
- network intrusion detection
- co training
- text categorization
- anomalous behavior
- active learning
- learning problems
- labeled and unlabeled data
- class labels
- multi label
- training examples
- pairwise
- labeled instances
- unlabeled instances
- one class support vector machines
- label information
- pairwise constraints
- neural network
- text mining
- natural language processing
- training set
- object recognition
- detect anomalies
- image processing