One-Shot Supervised Reinforcement Learning for Multi-targeted Tasks: RL-SAS.
Johane TakeuchiHiroshi TsujinoPublished in: ICANN (2) (2010)
Keyphrases
- reinforcement learning
- learning algorithm
- transfer learning
- supervised learning
- function approximation
- reinforcement learning algorithms
- state space
- semi supervised
- multi agent environments
- optimal policy
- model free
- markov decision processes
- learning agents
- multi agent
- temporal difference learning
- action selection
- policy search
- complex domains
- real robot
- reinforcement learning methods
- partially observable domains
- reinforcement learning agents
- temporal difference
- direct policy search
- dynamic programming
- unsupervised learning
- rl algorithms
- autonomous learning
- solve complex tasks
- learning capabilities
- active learning
- state and action spaces
- learning problems
- actor critic
- multi agent reinforcement learning
- evaluation function
- robotic systems
- supervised classification
- average reward
- function approximators
- action space
- policy iteration
- markov decision process
- control problems