Markov Decision Processes on Borel Spaces with Total Cost and Random Horizon.
Hugo Cruz-SuárezRocio Ilhuicatzi-RoldánRaúl Montes-de-OcaPublished in: J. Optim. Theory Appl. (2014)
Keyphrases
- markov decision processes
- total cost
- finite state
- optimal policy
- average cost
- reinforcement learning
- discount factor
- state space
- dynamic programming
- policy iteration
- optimal solution
- transition matrices
- service level
- minimum total cost
- finite horizon
- production cost
- decision theoretic planning
- markov decision process
- planning horizon
- inventory level
- holding cost
- learning algorithm
- partially observable
- reward function
- average reward
- markov decision problems
- risk sensitive
- infinite horizon
- action space
- search space