A Small Sphere and Large Margin Approach for Novelty Detection Using Training Data with Outliers.
Mingrui WuJieping YePublished in: IEEE Trans. Pattern Anal. Mach. Intell. (2009)
Keyphrases
- novelty detection
- training data
- text filtering
- anomaly detection
- support vector data description
- concept drift
- sentence level
- supervised learning
- learning algorithm
- minimum volume
- training set
- labeled data for training
- noisy data
- training samples
- feature vectors
- decision trees
- unlabeled data
- training examples
- labeled data
- input data
- small number
- prior knowledge
- data analysis
- neural network