Geodesic Gaussian kernels for value function approximation.
Masashi SugiyamaHirotaka HachiyaChristopher TowellSethu VijayakumarPublished in: Auton. Robots (2008)
Keyphrases
- gaussian kernels
- learning rate
- kernel function
- basis functions
- gaussian kernel
- high dimensional feature space
- distance transform
- reproducing kernel hilbert space
- state space
- shortest path
- feature space
- kernel principal component analysis
- geodesic distance
- kernel learning
- least squares
- machine learning
- convergence rate
- multiple kernel learning
- kernel methods
- data points
- pairwise
- reinforcement learning