Bayes-Adaptive Monte-Carlo Planning for Type-Based Reasoning in Large Partially Observable, Multi-Agent Environments.
Jonathon SchwartzHanna KurniawatiPublished in: AAMAS (2023)
Keyphrases
- monte carlo
- partially observable
- multi agent environments
- reinforcement learning
- state space
- decision problems
- markov chain
- dynamical systems
- markov decision problems
- markov decision processes
- reinforcement learning algorithms
- multi agent
- planning domains
- belief state
- temporal difference
- particle filter
- infinite horizon
- autonomous agents
- single agent
- random walk
- reward function
- heuristic search
- learning capabilities
- orders of magnitude
- multi agent systems
- control problems
- function approximation
- reinforcement learning methods
- planning problems
- learning algorithm