Feature Selection for classification of hyperspectral data by minimizing a tight bound on the VC dimension.
Phool PreetSanjit Singh Batra JayadevaPublished in: CoRR (2015)
Keyphrases
- vc dimension
- hyperspectral data
- upper bound
- feature selection
- lower bound
- hyperspectral images
- hyperspectral
- hyperspectral imagery
- classification accuracy
- worst case
- sample size
- multispectral
- support vector
- random projections
- support vector machine
- machine learning
- feature extraction
- model selection
- generalization bounds
- feature space
- sample complexity
- text classification
- pattern recognition
- remote sensing
- decision trees
- inductive inference
- feature set
- infrared
- high dimensionality
- feature vectors
- principal components
- generalization error
- text categorization
- supervised learning
- np hard
- image analysis
- image segmentation
- image processing
- learning algorithm