Acquisition by robots of danger-avoidance behaviors using probability-based reinforcement learning.
Daiki TakeyamaMasayoshi KanohTohgoroh MatsuiTsuyoshi NakamuraPublished in: FUZZ-IEEE (2015)
Keyphrases
- real robot
- reinforcement learning
- mobile robot
- multi robot
- robotic control
- robot control
- robotic systems
- function approximation
- unknown environments
- probability distribution
- cooperative
- humanoid robot
- robot behavior
- cooperative behavior
- imitation learning
- human robot interaction
- autonomous robots
- state space
- robot soccer
- learning algorithm
- optimal policy
- transfer learning
- learning agents
- optimal control
- markov decision processes
- data acquisition
- action selection
- human behavior
- acquisition process
- action space
- reinforcement learning algorithms
- model free
- manipulation tasks
- evolutionary learning
- multi agent reinforcement learning
- conditional probabilities
- machine learning
- control system
- robotic agents