Semi-supervised novelty detection using ensembles with regularized disagreement.
Alexandru TifreaEric StavaracheFanny YangPublished in: UAI (2022)
Keyphrases
- novelty detection
- semi supervised
- active learning
- text filtering
- semi supervised learning
- anomaly detection
- ensemble learning
- unsupervised learning
- unlabeled data
- labeled data
- concept drift
- pairwise
- supervised learning
- pairwise constraints
- decision trees
- metric learning
- risk minimization
- support vector data description
- labeled and unlabeled data
- ensemble methods
- training examples
- neural network
- pattern recognition
- machine learning
- base classifiers
- loss function
- generalization error
- non stationary
- expectation maximization
- probabilistic model