Simplified Belief-Dependent Reward MCTS Planning with Guaranteed Tree Consistency.
Ori SztyglicAndrey ZhitnikovVadim IndelmanPublished in: CoRR (2021)
Keyphrases
- tree structure
- belief space
- belief state
- belief revision
- motion planning
- reinforcement learning
- ai planning
- tree structures
- long run
- domain independent
- planning problems
- tree search
- heuristic search
- monte carlo tree search
- global constraints
- spanning tree
- expected utility
- binary tree
- global consistency
- tree nodes
- convex constraints