Effect of human guidance and state space size on Interactive Reinforcement Learning.
Halit Bener SuaySonia ChernovaPublished in: RO-MAN (2011)
Keyphrases
- state space
- reinforcement learning
- reinforcement learning algorithms
- markov decision processes
- optimal policy
- dynamic programming
- function approximation
- action space
- markov decision process
- markov chain
- human subjects
- heuristic search
- temporal difference
- learning algorithm
- planning problems
- state variables
- belief state
- user interaction
- dynamical systems
- human interaction
- model free
- particle filter
- user friendly
- partially observable
- initial state
- learning agent
- search space
- guided exploration
- state abstraction
- multi agent
- computational complexity
- virtual humans
- computer graphics
- decision problems