Reinforcement learning for high-dimensional problems with symmetrical actions.
Md. Abdus Samad KamalJunichi MurataPublished in: SMC (7) (2004)
Keyphrases
- high dimensional problems
- reinforcement learning
- action selection
- perceptual aliasing
- dimension reduction
- high dimensional
- action space
- partially observable
- state and action spaces
- state space
- multiagent reinforcement learning
- reinforcement learning algorithms
- markov decision processes
- learning agent
- state action
- partial observability
- partially observable domains
- decision theoretic
- function approximation
- reward function
- machine learning
- learning algorithm
- policy search
- model free
- computer vision
- sensory inputs
- high dimensionality
- optimal policy
- probabilistic model
- learning process
- training set
- preprocessing
- reward signal
- agent receives
- behavioural cloning