Reinforcement Learning with Parameterized Actions.
Warwick MassonPravesh RanchodGeorge Dimitri KonidarisPublished in: AAAI (2016)
Keyphrases
- reinforcement learning
- action selection
- perceptual aliasing
- action space
- partially observable
- function approximation
- state and action spaces
- reward function
- partially observable domains
- state space
- markov decision processes
- multi agent
- state action
- multiagent reinforcement learning
- reinforcement learning algorithms
- partial observability
- model free
- reinforcement learning methods
- action sets
- machine learning
- previously learned
- behavioural cloning
- reasoning about actions
- learning agent
- plan recognition
- transfer learning
- optimal policy
- learning process
- learning algorithm
- markov decision process
- learned knowledge
- decision theoretic
- learning problems
- continuous state
- partially observable markov decision process
- initially unknown
- human activities
- mobile robot
- robotic control