Discounted cost optimality problem: stability with respect to weak metrics.
Evgueni GordienkoEnrique Lemus-RodríguezRaúl Montes-de-OcaPublished in: Math. Methods Oper. Res. (2008)
Keyphrases
- average cost
- markov decision processes
- finite number
- infinite horizon
- long run
- total cost
- optimal policy
- finite horizon
- linear programming
- optimal control
- high cost
- average reward
- multistage
- finite state
- similarity metrics
- stability analysis
- information retrieval
- linear program
- production cost
- cash flow
- expected cost
- quality measures
- multi class
- reinforcement learning
- learning algorithm