Gradient-based kernel method for feature extraction and variable selection.
Kenji FukumizuChenlei LengPublished in: NIPS (2012)
Keyphrases
- variable selection
- kernel methods
- feature extraction
- dimension reduction
- feature space
- high dimensional
- input variables
- kernel principal component analysis
- cross validation
- feature selection
- support vector
- kernel matrix
- preprocessing
- learning problems
- kernel function
- machine learning
- support vector machine
- principal component analysis
- kernel pca
- dimensionality reduction
- model selection
- image processing
- feature vectors
- linear discriminant analysis
- reproducing kernel hilbert space
- face recognition
- discriminant analysis
- learning tasks
- pattern recognition
- high dimensionality
- low dimensional
- reproducing kernel
- classification accuracy
- high dimensional data
- ls svm
- multiple kernel learning
- cluster analysis
- singular value decomposition
- input space
- clustering algorithm
- support vector machine svm
- learning algorithm
- unsupervised learning
- neural network