Error bounds for suboptimal solutions to kernel principal component analysis.
Giorgio GneccoMarcello SanguinetiPublished in: Optim. Lett. (2010)
Keyphrases
- error bounds
- kernel principal component analysis
- suboptimal solutions
- metaheuristic
- search algorithm
- optimal solution
- kernel pca
- feature extraction
- discriminant analysis
- upper bound
- principal components
- worst case
- theoretical analysis
- principal component analysis
- kernel function
- preprocessing
- feature space
- classification method
- high dimensional
- kernel methods
- face recognition
- tuning parameters
- support vector machine svm
- search procedures
- feature vectors
- kernel matrix
- gabor wavelets
- np complete problems
- optimization problems
- feature selection
- linear discriminant analysis
- np hard
- learning algorithm
- genetic algorithm
- simulated annealing
- metric learning
- graph kernels
- high dimensional feature space
- lower bound