Will My Robot Achieve My Goals? Predicting the Probability that an MDP Policy Reaches a User-Specified Behavior Target.
Alexander GuyerThomas G. DietterichPublished in: CoRR (2022)
Keyphrases
- optimal policy
- markov decision process
- human friendly
- user interaction
- mobile robot
- autonomous robots
- user interface
- end users
- vision system
- markov decision processes
- relevance feedback
- probability distribution
- home environment
- robot behavior
- behavioral patterns
- dynamic programming
- partially observable
- human robot interaction
- human users
- path planning
- user preferences
- user profiles
- markov decision problems
- reinforcement learning
- unknown environments
- visual servoing
- robot navigation
- infinite horizon
- utility function