Sketching, Embedding and Dimensionality Reduction in Information Theoretic Spaces.
Amirali AbdullahRavi KumarAndrew McGregorSergei VassilvitskiiSuresh VenkatasubramanianPublished in: AISTATS (2016)
Keyphrases
- information theoretic
- dimensionality reduction
- embedding space
- low dimensional spaces
- latent space
- nonlinear dimensionality reduction
- mutual information
- structure preserving
- graph embedding
- information theory
- low dimensional
- manifold learning
- high dimensional
- multidimensional scaling
- theoretic framework
- principal component analysis
- high dimensional data
- euclidean space
- input space
- jensen shannon divergence
- information bottleneck
- data representation
- feature selection
- data points
- information theoretic measures
- feature extraction
- pattern recognition
- entropy measure
- log likelihood
- euclidean distance
- kullback leibler divergence
- vector space
- minimum description length
- image processing
- riemannian manifolds
- dimensionality reduction methods
- feature space
- geometric structure
- relative entropy
- multi modality
- metric learning
- distributional clustering