Reusable Reinforcement Learning for Modular Self Motivated Agents.
Jaroslav VitkuPavel NahodilPublished in: ECMS (2014)
Keyphrases
- reinforcement learning
- multi agent
- learning agents
- multi agent systems
- agent receives
- multiagent systems
- multi agent reinforcement learning
- action selection
- multiple agents
- intelligent agents
- single agent
- software agents
- multi agent environments
- cooperative
- multiagent learning
- reinforcement learning agents
- autonomous agents
- learning agent
- agent architecture
- learning capabilities
- agent behavior
- dynamic environments
- robocup soccer
- state space
- mobile agents
- function approximation
- model free
- optimal policy
- multiagent reinforcement learning
- partial observability
- agent interactions
- artificial agents
- distributed artificial intelligence
- optimal control
- heterogeneous agents
- interacting agents
- coalition formation
- learned knowledge
- incomplete information
- game theory
- state abstraction
- decentralized control
- software systems
- agent systems
- reinforcement learning algorithms
- decision making