Stochastic kernel temporal difference for reinforcement learning.
Jihye BaeLuis G. Sánchez GiraldoPratik ChhatbarJoseph T. FrancisJustin C. SanchezJosé C. PríncipePublished in: MLSP (2011)
Keyphrases
- temporal difference
- reinforcement learning
- monte carlo
- function approximation
- td learning
- reinforcement learning algorithms
- model free
- temporal difference learning
- evaluation function
- action selection
- function approximators
- kernel methods
- kernel function
- step size
- temporal difference methods
- state space
- policy evaluation
- policy iteration
- actor critic
- support vector
- learning algorithm
- feature space
- optimal policy
- markov chain
- active learning
- learning process
- stochastic processes
- continuous state
- policy search
- reinforcement learning problems
- td methods