Sparse Hilbert Schmidt Independence Criterion and Surrogate-Kernel-Based Feature Selection for Hyperspectral Image Classification.
Bharath Bhushan DamodaranNicolas CourtySébastien LefèvrePublished in: IEEE Trans. Geosci. Remote. Sens. (2017)
Keyphrases
- hilbert schmidt independence criterion
- feature selection
- hyperspectral image classification
- active learning
- hyperspectral images
- feature space
- hyperspectral
- text classification
- mutual information
- kernel learning
- classification accuracy
- feature set
- dimensionality reduction
- high dimensional
- support vector
- feature extraction
- machine learning
- support vector machine
- multi class
- knn
- hyperspectral data
- model selection
- sparse representation
- kernel methods
- high dimensionality
- multi task
- random projections
- pattern recognition
- learning algorithm