On discovery and learning of models with predictive representations of state for agents with continuous actions and observations.
David WingateSatinder P. SinghPublished in: AAMAS (2007)
Keyphrases
- state transitions
- multi agent
- learning agents
- learning algorithm
- learned knowledge
- learning models
- learning process
- prior knowledge
- action sequences
- decision theoretic
- action selection
- accurate models
- agent behavior
- multi agent systems
- predictive state representations
- learning mechanisms
- rational agents
- multiple agents
- intelligent behavior
- multiagent reinforcement learning
- automatically discovering
- learning capabilities
- partially observable
- cooperative
- multiple representations
- past observations
- state action
- learning agent
- qualitative models
- incomplete information
- software agents
- autonomous agents
- intelligent agents
- probabilistic model
- reinforcement learning