Parameter Tuning via Kernel Matrix Approximation for Support Vector Machine.
Chenhao YangLizhong DingShizhong LiaoPublished in: J. Comput. (2012)
Keyphrases
- parameter tuning
- support vector machine
- matrix approximation
- kernel function
- kernel methods
- support vector
- least squares
- theoretical guarantees
- approximation error
- feature space
- maximum entropy
- svm classifier
- multi class
- machine learning
- bregman divergences
- parameter settings
- multiple kernel learning
- training data
- feature selection
- input space
- low rank matrix approximation
- cost sensitive
- kernel matrix
- neural network
- hyperplane
- kernel machines
- binary classification
- knn
- radial basis function
- model selection