Acceleration of Reinforcement Learning by Policy Evaluation Using Nonstationary Iterative Method.
Kei SendaSuguru HattoriToru HishinumaTakehisa KohdaPublished in: IEEE Trans. Cybern. (2014)
Keyphrases
- non stationary
- policy evaluation
- reinforcement learning
- temporal difference
- least squares
- model free
- function approximation
- monte carlo
- policy iteration
- markov decision processes
- td learning
- variance reduction
- state space
- reinforcement learning algorithms
- random fields
- machine learning
- semi parametric
- partially observable markov decision processes
- evaluation function
- optimal policy
- learning algorithm
- step size
- supervised learning
- transfer learning
- dynamic programming
- computational complexity
- markov decision problems
- multi agent