Provably Efficient Reinforcement Learning in Partially Observable Dynamical Systems.
Masatoshi UeharaAyush SekhariJason D. LeeNathan KallusWen SunPublished in: NeurIPS (2022)
Keyphrases
- partially observable
- dynamical systems
- reinforcement learning
- state space
- partial observability
- hidden state
- partially observable domains
- markov decision problems
- dynamic systems
- partially observable environments
- decision problems
- markov decision processes
- partially observable markov decision process
- action models
- partial observations
- nonlinear dynamical systems
- partially observable markov decision processes
- machine learning
- objective function
- dynamic programming
- control system