A Pairwise Covariance-Preserving Projection Method for Dimension Reduction.
Xiaoming LiuZhaohui WangZhilin FengJinshan TangPublished in: ICDM (2007)
Keyphrases
- dimension reduction
- projection method
- pairwise
- harmonic functions
- principal component analysis
- feature extraction
- high dimensional
- linear projection
- linear discriminant analysis
- high dimensional problems
- high dimensional data
- random projections
- low dimensional
- data mining and machine learning
- variable selection
- loss function
- discriminative information
- semi supervised
- unsupervised learning
- feature space
- high dimensional data analysis
- singular value decomposition
- metric learning
- manifold learning
- high dimensionality
- cluster analysis
- feature selection
- partial least squares
- preprocessing
- covariance matrix
- similarity measure
- dimensionality reduction
- dimension reduction methods
- spectral clustering
- data sets
- face recognition
- data mining