Dimensionality Reduction has Quantifiable Imperfections: Two Geometric Bounds.
Kry Yik Chau LuiGavin Weiguang DingRuitong HuangRobert J. McCannPublished in: NeurIPS (2018)
Keyphrases
- dimensionality reduction
- pattern recognition
- high dimensionality
- principal component analysis
- low dimensional
- high dimensional
- pattern recognition and machine learning
- lower bound
- feature extraction
- linear discriminant analysis
- geometric information
- data representation
- high dimensional data
- upper bound
- nonlinear dimensionality reduction
- dimensionality reduction methods
- structure preserving
- error bounds
- principal components
- geometric structure
- feature selection
- computer vision
- kernel learning
- manifold learning
- input space
- upper and lower bounds
- neural network
- lower dimensional
- lower and upper bounds
- learning algorithm
- sparse representation
- worst case
- data points
- vc dimension
- contingency tables
- support vector machine
- linear dimensionality reduction