Ensembles for Continuous Actions in Reinforcement Learning.
Siegmund DuellSteffen UdluftPublished in: ESANN (2013)
Keyphrases
- action space
- reinforcement learning
- action selection
- state space
- state and action spaces
- perceptual aliasing
- continuous state
- continuous action
- markov decision processes
- real valued
- partially observable
- continuous state spaces
- reinforcement learning algorithms
- function approximation
- multiagent reinforcement learning
- state action
- partially observable domains
- reasoning about actions
- continuous state and action spaces
- plan recognition
- ensemble learning
- learning agent
- continuous domains
- reward function
- transfer learning
- multi agent
- decision trees
- machine learning
- markov decision process
- learned knowledge
- reinforcement learning methods
- random forests
- partially observable markov decision process
- learning process
- learning algorithm