Single believe state generation for partially observable real-time strategy games.
Alberto UriarteSantiago OntañónPublished in: CIG (2017)
Keyphrases
- partially observable
- state space
- real time strategy games
- belief state
- partial observations
- reinforcement learning
- decision problems
- partial observability
- markov decision processes
- dynamical systems
- hidden state
- fully observable
- reward function
- markov decision problems
- evaluation function
- computational complexity
- initially unknown
- objective function