Tradeoff between exploration and exploitation of OQ(lambda) with non-Markovian update in dynamic environments.
Maryam ShokriHamid R. TizhooshMohamed S. KamelPublished in: IJCNN (2008)
Keyphrases
- dynamic environments
- reinforcement learning agents
- mobile robot
- autonomous agents
- reinforcement learning
- path planning
- search capabilities
- exploration exploitation tradeoff
- decision processes
- agent systems
- situation calculus
- collision avoidance
- fixed point
- single agent
- changing environment
- action selection
- real environment
- scheduling strategy
- transfer learning
- stochastic process
- scheduling algorithm
- computational complexity
- potential field
- search algorithm
- visual slam
- highly dynamic environments
- learning algorithm