Unsupervised discovery of co-occurrence in sparse high dimensional data.
Ondrej ChumJiri MatasPublished in: CVPR (2010)
Keyphrases
- locally linear embedding
- co occurrence
- high dimensional data
- high dimensional
- sparse representation
- dimensionality reduction
- low dimensional
- underlying manifold
- nearest neighbor
- high dimensionality
- manifold learning
- subspace clustering
- similarity search
- dimension reduction
- data points
- wordnet
- data sets
- data analysis
- high dimensional spaces
- semantic relations
- high dimensional datasets
- low rank
- named entities
- visual words
- text data
- principal component analysis
- multiple types
- linear discriminant analysis
- feature extraction
- latent semantic analysis
- database
- word co occurrence
- topic models
- information retrieval systems
- information extraction
- knn
- image data
- feature space