A novel dimensionality reduction technique based on kernel optimization through graph embedding.
Nicholas VretosAnastasios TefasIoannis PitasPublished in: Signal Image Video Process. (2015)
Keyphrases
- graph embedding
- low dimensional
- dimensionality reduction
- semi supervised
- data representation
- discriminant analysis
- high dimensional
- high dimensional data
- subspace learning
- discriminant embedding
- data points
- feature space
- dimension reduction
- feature representation
- kernel machines
- neural network
- principal component analysis
- pattern recognition
- feature extraction
- geometric properties
- dictionary learning
- decision trees
- data sets