Outlier preservation by dimensionality reduction techniques.
Martijn OnderwaterPublished in: Int. J. Data Anal. Tech. Strateg. (2015)
Keyphrases
- dimensionality reduction
- outlier detection
- low dimensional
- high dimensionality
- high dimensional data
- high dimensional
- principal component analysis
- data representation
- random projections
- dimensionality reduction methods
- pattern recognition
- feature selection
- data points
- pattern recognition and machine learning
- manifold learning
- linear discriminant analysis
- structure preserving
- singular value decomposition
- feature extraction
- feature space
- euclidean distance
- principal components analysis
- principal components
- metric learning
- lower dimensional
- multidimensional scaling
- nonlinear dimensionality reduction
- linear dimensionality reduction
- intrinsic dimensionality
- novelty detection
- discriminant analysis
- image processing
- preprocessing step
- kernel pca
- sparse representation
- digital preservation
- input space
- dimension reduction
- data sets
- active learning
- data streams
- data mining
- linear projection
- outlier mining
- databases