Sparse Kernel-Based Least Squares Temporal Difference with Prioritized Sweeping.
Cijia SunXinghong LingYuchen FuQuan LiuHaijun ZhuJianwei ZhaiPeng ZhangPublished in: ICONIP (3) (2016)
Keyphrases
- temporal difference
- least squares
- policy evaluation
- sparse linear
- reinforcement learning
- td learning
- evaluation function
- function approximation
- monte carlo
- policy iteration
- model free
- step size
- action selection
- reinforcement learning algorithms
- temporal difference methods
- kernel methods
- high dimensional
- supervised learning
- machine learning
- optical flow
- function approximators
- cost function
- support vector
- feature vectors