Reinforcement Learning in Feature Space: Matrix Bandit, Kernels, and Regret Bound.
Lin YangMengdi WangPublished in: ICML (2020)
Keyphrases
- regret bounds
- feature space
- reinforcement learning
- dot product
- kernel function
- lower bound
- online learning
- linear regression
- upper bound
- feature vectors
- kernel methods
- mean shift
- principal component analysis
- feature selection
- support vector machine
- feature set
- state space
- classification accuracy
- dimensionality reduction
- training samples
- data points
- high dimensional
- learning algorithm
- least squares
- linear combination
- markov decision processes
- kernel matrix
- bregman divergences
- e learning