Graph-based transformation manifolds for invariant pattern recognition with kernel methods.
Alexei PozdnoukhovSamy BengioPublished in: ICPR (4) (2006)
Keyphrases
- kernel methods
- pattern recognition
- machine learning
- feature space
- learning problems
- kernel function
- support vector
- kernel matrix
- manifold learning
- image processing
- support vector machine
- feature extraction
- dimensionality reduction
- support vector machine svm
- multiple kernel learning
- computer vision
- semi supervised
- low dimensional
- high dimensional
- kernel pca
- kernel trick
- geometric structure
- neural network
- reproducing kernel hilbert space
- multiple kernel
- kernel matrices
- input space
- euclidean space
- graph kernels
- kernel ridge regression
- kernel fisher discriminant analysis