An insect-based method for learning landmark reliability using expectation reinforcement in dynamic environments.
Zenon MathewsPaul F. M. J. VerschureSergi Bermúdez i BadiaPublished in: ICRA (2010)
Keyphrases
- dynamic environments
- reinforcement learning
- prior knowledge
- learning scheme
- neural network
- unsupervised learning
- learning algorithm
- potential field
- learning process
- mobile robot
- dynamic programming
- significant improvement
- objective function
- supervised learning
- detection method
- autonomous agents
- agent systems
- motor control