Multi-Task Kernel Null-Space for One-Class Classification.
Shervin Rahimzadeh ArashlooJosef KittlerPublished in: CoRR (2019)
Keyphrases
- multi task
- null space
- novelty detection
- learning tasks
- linear discriminant analysis
- kernel methods
- metric learning
- singular value decomposition
- learning problems
- feature space
- transfer learning
- feature selection
- kernel function
- multi class
- discriminative information
- kernel matrix
- principal components
- machine learning
- reproducing kernel hilbert space
- discriminant analysis
- dimension reduction
- information gain
- support vector
- singular values
- theoretical guarantees
- pairwise
- feature vectors