Near-Optimal Time and Sample Complexities for Solving Markov Decision Processes with a Generative Model.
Aaron SidfordMengdi WangXian WuLin YangYinyu YePublished in: NeurIPS (2018)
Keyphrases
- generative model
- markov decision processes
- transition matrices
- semi markov decision processes
- probabilistic model
- optimal policy
- state space
- reinforcement learning
- finite state
- markov decision problems
- policy iteration
- dynamic programming
- decision theoretic planning
- posterior probability
- prior knowledge
- average reward
- reward function
- partially observable
- bayesian framework
- model based reinforcement learning
- reachability analysis
- em algorithm
- average cost
- markov decision process
- action space
- planning under uncertainty
- factored mdps
- stochastic shortest path
- infinite horizon
- action sets
- hidden variables
- topic models
- expectation maximization
- linear programming
- search algorithm