Recurrence Conditions for Average and Blackwell Optimality in Denumerable State Markov Decision Chains.
Rommert DekkerArie HordijkPublished in: Math. Oper. Res. (1992)
Keyphrases
- markov decision chains
- average cost
- markov decision processes
- long run
- initial state
- finite state
- optimal policy
- risk sensitive
- state space
- sufficient conditions
- infinite horizon
- linear program
- control policy
- finite horizon
- markov decision process
- finite number
- total cost
- optimal control
- multistage
- markov chain
- stationary policies
- machine learning
- steady state
- linear programming
- decision making
- dynamic programming
- control system
- lower bound
- computational complexity