Modelling Policies in MDPs in Reproducing Kernel Hilbert Space.
Guy LeverRonnie StaffordPublished in: AISTATS (2015)
Keyphrases
- reproducing kernel hilbert space
- optimal policy
- markov decision processes
- kernel methods
- loss function
- learning theory
- euclidean space
- kernel function
- special case
- reinforcement learning
- data dependent
- state space
- density estimation
- real valued
- input space
- learning problems
- domain adaptation
- gaussian process
- distance measure
- kernel matrix
- support vector
- linear model
- machine learning
- pairwise
- similarity measure
- similarity search
- least squares
- knn
- feature extraction