Optimal learning rates for Kernel Conjugate Gradient regression.
Gilles BlanchardNicole KrämerPublished in: NIPS (2010)
Keyphrases
- conjugate gradient
- reproducing kernel hilbert space
- gaussian kernels
- learning rate
- convergence rate
- kernel methods
- training algorithm
- loss function
- kernel function
- special case
- density estimation
- learning theory
- learning algorithm
- learning problems
- real valued
- support vector
- kernel matrix
- input space
- gaussian process
- positive definite
- data dependent
- euclidean space
- optimal solution
- feature space
- distance measure
- levenberg marquardt
- convergence speed
- support vector machine
- regression model
- linear model
- domain adaptation
- faster convergence
- model selection
- dimensionality reduction
- multiple kernel learning
- genetic algorithm
- hidden layer
- support vector regression