Solving Markov decision processes for network-level post-hazard recovery via simulation optimization and rollout.
Yugandhar SarkaleSaeed NozhatiEdwin K. P. ChongBruce R. EllingwoodHussam MahmoudPublished in: CoRR (2018)
Keyphrases
- markov decision processes
- transition matrices
- semi markov decision processes
- finite state
- state space
- reinforcement learning
- optimal policy
- planning under uncertainty
- decision theoretic planning
- policy iteration
- dynamic programming
- reachability analysis
- factored mdps
- markov decision problems
- average reward
- model based reinforcement learning
- stochastic shortest path
- decision processes
- finite horizon
- average cost
- action space
- markov decision process
- partially observable
- infinite horizon
- probabilistic planning
- risk sensitive
- state abstraction
- reinforcement learning algorithms
- constraint satisfaction problems
- real time dynamic programming
- machine learning