Regret based Robust Solutions for Uncertain Markov Decision Processes.
Asrar AhmedPradeep VarakanthamYossiri AdulyasakPatrick JailletPublished in: NIPS (2013)
Keyphrases
- markov decision processes
- total reward
- reward function
- optimal policy
- expected reward
- transition matrices
- reinforcement learning
- finite state
- state space
- dynamic programming
- average reward
- decision theoretic planning
- decision making
- policy iteration
- finite horizon
- reinforcement learning algorithms
- planning under uncertainty
- risk sensitive
- average cost
- markov decision process
- loss function
- model based reinforcement learning
- factored mdps
- partially observable
- infinite horizon
- partially observed
- reachability analysis
- probabilistic planning
- decision processes
- lower bound
- linear programming
- markov chain