Partially Observable Total-Cost Markov Decision Processes with Weakly Continuous Transition Probabilities.
Eugene A. FeinbergPavlo O. KasyanovMichael Z. ZgurovskyPublished in: Math. Oper. Res. (2016)
Keyphrases
- total cost
- partially observable
- markov decision processes
- transition probabilities
- reward function
- markov decision problems
- markov chain
- action space
- state space
- markov models
- random walk
- finite state
- average cost
- reinforcement learning
- optimal policy
- optimal solution
- policy iteration
- reinforcement learning algorithms
- markov decision process
- dynamic programming
- infinite horizon
- average reward
- inventory level
- finite horizon
- maximum entropy
- partially observable markov decision processes
- hidden markov models
- np hard
- dynamical systems
- decision problems
- lead time
- multiple agents
- hidden variables
- decision processes
- data mining
- partially observable markov decision process