PAC Reinforcement Learning for Predictive State Representations.
Wenhao ZhanMasatoshi UeharaWen SunJason D. LeePublished in: ICLR (2023)
Keyphrases
- predictive state representations
- reinforcement learning
- temporal difference
- dynamical systems
- function approximation
- partially observable markov decision processes
- model free
- state space
- stochastic systems
- learning algorithm
- evaluation function
- past observations
- partially observable
- step size
- monte carlo
- sample size
- optimal policy
- supervised learning
- multi agent
- machine learning
- decision problems
- markov decision processes