Application of Newton's Method to action selection in continuous state- and action-space reinforcement learning.
Barry D. NicholsDimitris C. DracopoulosPublished in: ESANN (2014)
Keyphrases
- action space
- reinforcement learning
- continuous state
- action selection
- continuous state and action spaces
- continuous state spaces
- policy search
- state space
- dynamic programming
- real valued
- state action
- model free
- markov decision processes
- control policies
- stochastic processes
- function approximators
- computational complexity
- data mining
- hidden markov models