Locally Minimizing Embedding and Globally Maximizing Variance: Unsupervised Linear Difference Projection for Dimensionality Reduction.
Minghua WanZhihui LaiZhong JinPublished in: Neural Process. Lett. (2011)
Keyphrases
- dimensionality reduction
- discriminant projection
- nonlinear dimensionality reduction
- unsupervised learning
- structure preserving
- graph embedding
- locality preserving projections
- linear projection
- data representation
- maximum variance
- dimensionality reduction methods
- principal component analysis
- embedding space
- manifold learning
- high dimensional data
- multidimensional scaling
- feature extraction
- semi supervised
- high dimensional
- neighborhood preserving
- linear discriminant analysis
- principal components
- convex functions
- pattern recognition
- low dimensional
- random projections
- low dimensional spaces
- high dimensionality
- discriminant analysis
- feature space
- active learning
- dimension reduction
- locally linear embedding
- input space
- subspace learning
- convex sets
- piecewise linear
- prediction error
- singular value decomposition
- output space
- vector space
- euclidean distance
- linear model
- geodesic distance