Dimension and Margin Bounds for Reflection-invariant Kernels.
Thorsten DoliwaMichael KallweitHans Ulrich SimonPublished in: COLT (2008)
Keyphrases
- support vector
- lower bound
- half spaces
- upper bound
- translation invariant
- affine invariant
- kernel methods
- upper and lower bounds
- affine transformation
- feature space
- maximum margin
- margin maximization
- kernel function
- lower and upper bounds
- rademacher complexity
- moment invariants
- error bounds
- worst case
- learning machines
- large margin classifiers
- training set
- decision boundary
- multiple kernel learning
- invariant features
- linear combination
- generalization bounds
- multi class
- data sets