Optimally solving Markov decision processes with total expected discounted reward function: Linear programming revisited.
Oguzhan AlagözMehmet U. S. AyvaciJeffrey T. LinderothPublished in: Comput. Ind. Eng. (2015)
Keyphrases
- markov decision processes
- reward function
- linear programming
- dynamic programming
- transition matrices
- markov decision problems
- semi markov decision processes
- reinforcement learning algorithms
- optimal policy
- state space
- partially observable
- policy iteration
- finite state
- reinforcement learning
- markov decision process
- average reward
- inverse reinforcement learning
- decision theoretic planning
- factored mdps
- stochastic shortest path
- average cost
- infinite horizon
- linear program
- finite horizon
- state and action spaces
- action space
- decision problems
- sufficient conditions
- hierarchical reinforcement learning
- lower bound
- objective function
- partially observable markov decision processes
- transition model
- hidden markov models