Online Markov decision processes with non-oblivious strategic adversary.
Le Cong DinhDavid Henry MguniLong Tran-ThanhJun WangYaodong YangPublished in: Auton. Agents Multi Agent Syst. (2023)
Keyphrases
- markov decision processes
- optimal policy
- finite state
- transition matrices
- state space
- reachability analysis
- finite horizon
- reinforcement learning
- dynamic programming
- reinforcement learning algorithms
- policy iteration
- planning under uncertainty
- risk sensitive
- factored mdps
- average cost
- markov decision process
- partially observable
- decision theoretic planning
- model based reinforcement learning
- infinite horizon
- decision making
- action space
- stochastic shortest path
- sufficient conditions
- semi markov decision processes
- utility function
- action sets
- state abstraction
- average reward