Kernel principal subspace Mahalanobis distances for outlier detection.
Cong LiMichael GeorgiopoulosGeorgios C. AnagnostopoulosPublished in: IJCNN (2011)
Keyphrases
- outlier detection
- distance function
- euclidean distance
- feature space
- distance measure
- detection algorithm
- density ratio estimation
- knowledge discovery
- fraud detection
- distance metric
- kernel methods
- dimensionality reduction
- high dimensional datasets
- kernel function
- density estimation
- metric learning
- detect outliers
- low dimensional
- high dimensional
- detecting outliers
- credit card fraud detection
- data mining
- high dimensional data
- data streams
- support vector
- kernel matrix
- distance based outlier detection
- nearest neighbor
- feature extraction
- kernel density estimation
- similarity search
- video sequences
- databases