A unified algorithm framework for mean-variance optimization in discounted Markov decision processes.
Shuai MaXiaoteng MaLi XiaPublished in: CoRR (2022)
Keyphrases
- markov decision processes
- dynamic programming
- average reward
- optimal policy
- model based reinforcement learning
- reinforcement learning
- search space
- computational complexity
- probabilistic model
- state space
- policy iteration
- learning algorithm
- monte carlo
- decision theoretic planning
- linear programming
- finite state
- markov decision process
- search algorithm
- convergence rate
- planning under uncertainty
- machine learning