Randomized Belief-Space Replanning in Partially-Observable Continuous Spaces.
Kris K. HauserPublished in: WAFR (2010)
Keyphrases
- belief space
- partially observable
- belief state
- state space
- partial observability
- planning domains
- markov decision processes
- decision problems
- reinforcement learning
- dynamical systems
- infinite horizon
- classical planning
- markov decision problems
- heuristic search
- fully observable
- partially observable markov decision processes
- orders of magnitude
- motion planning
- planning problems
- planning under uncertainty
- dynamic environments
- decision theoretic
- dynamic programming
- reward function
- average cost
- domain specific
- learning algorithm