During the week of September 24, we will be implementing some features to improve ERA. Some users may encounter errors when depositing or editing files. We apologize for any potential inconvenience, and will remove this message when we are done our maintenance upgrade!
SearchSkip to Search Results
- 1Actor-critic reinforcement learning algorithms
- 1Approximate dynamic programming
- 1Function approximation
- 1Policy gradient methods
Technical report TR09-10. We present four new reinforcement learning algorithms based on actor-critic, function approximation, and natural gradient ideas, and we provide their convergence proofs. Actor-critic reinforcement learning methods are online approximations to policy iteration in which...