Global structure of policy search spaces for reinforcement learning.
Belinda StapelbergKatherine M. MalanPublished in: GECCO (Companion) (2019)
Keyphrases
- policy search
- global structure
- reinforcement learning
- reinforcement learning algorithms
- continuous state
- low dimensional
- geometric structure
- dynamic programming
- endpoints
- global information
- partially observable markov decision processes
- reward function
- function approximation
- policy gradient
- multi agent
- reinforcement learning methods
- wavelet coefficients
- dimensionality reduction
- multiscale
- state space
- temporal difference
- model free
- machine learning
- markov decision processes
- optimal policy
- markov decision process
- function approximators
- action selection
- markov decision problems
- tree structure