Policy Improvement Directions for Reinforcement Learning in Reproducing Kernel Hilbert Spaces.
Santiago PaternainJuan Andrés BazerqueAustin SmallAlejandro RibeiroPublished in: CDC (2019)
Keyphrases
- reproducing kernel hilbert space
- reinforcement learning
- optimal policy
- gaussian kernels
- reproducing kernel
- bayes risk
- uniform convergence
- loss function
- kernel methods
- learning problems
- special case
- covering numbers
- learning theory
- learning algorithm
- data dependent
- domain adaptation
- linear model
- kernel function
- state space
- euclidean space
- density estimation
- kernel matrix
- representer theorem
- machine learning
- gaussian process
- learning rate
- real valued
- hilbert space
- supervised learning
- feature space