Feature selection for SVM via optimization of kernel polarization with Gaussian ARD kernels.
Tinghua WangHoukuan HuangShengfeng TianJianfeng XuPublished in: Expert Syst. Appl. (2010)
Keyphrases
- rbf kernel
- support vector
- feature selection
- kernel function
- support vector machine
- polynomial kernels
- kernel machines
- gaussian kernel
- feature space
- radial basis function
- kernel methods
- multiple kernel learning
- optimal kernel
- kernel learning
- positive definite
- support vector machine svm
- svm classifier
- radial basis function kernel
- gaussian kernels
- multiple kernel
- kernel parameters
- svm classification
- representer theorem
- support vector classification
- reproducing kernel hilbert space
- classification accuracy
- hyperplane
- decision function
- string kernels
- machine learning
- support vector machines for classification
- feature extraction
- linear svm
- scale space
- support vectors
- standard svm
- learning machines
- text classification
- multi task
- binary classification
- feature vectors
- kernel matrix
- feature set
- text categorization
- dimensionality reduction
- multi class
- knn
- stochastic gradient descent
- tree kernels
- high dimensional
- positive semidefinite
- decision stumps
- high dimensional feature space
- support vector regression