Reinforcement learning with guided policy search using Gaussian processes.
Hunor JakabLehel CsatóPublished in: IJCNN (2012)
Keyphrases
- policy search
- gaussian processes
- reinforcement learning
- gaussian process
- continuous state
- reinforcement learning algorithms
- dynamic programming
- state space
- function approximation
- policy gradient
- markov decision problems
- markov decision processes
- model free
- pairwise
- reward function
- temporal difference
- hyperparameters
- multi agent
- action selection
- optimal control
- planning problems
- non stationary
- multi class
- support vector
- machine learning