L2, 0-norm regularization based feature selection for very high resolution remote sensing images.
Xi ChenYanfeng GuYe ZhangYiming YanPublished in: IGARSS (2015)
Keyphrases
- remote sensing images
- norm regularization
- feature selection
- convex optimization
- change detection
- remote sensing
- regularization methods
- multispectral
- canonical correlation analysis
- total variation
- matrix completion
- multiple kernel learning
- sparse representation
- group lasso
- loss function
- multi class
- satellite images
- text categorization
- low rank
- kernel learning
- max margin
- hyperspectral
- mutual information
- multi task
- variable selection
- support vector machine
- feature space
- support vector
- feature extraction
- incomplete data
- regression model
- image restoration
- text classification
- denoising
- high resolution
- solution path
- high quality
- machine learning
- multi task learning
- dimension reduction
- image denoising
- kernel function