A variable neighborhood search based algorithm for finite-horizon Markov Decision Processes.
Qiu-Hong ZhaoJack BrimbergNenad MladenovicPublished in: Appl. Math. Comput. (2010)
Keyphrases
- markov decision processes
- finite horizon
- dynamic programming
- optimal policy
- policy iteration
- computational complexity
- cost function
- objective function
- learning algorithm
- state space
- infinite horizon
- average cost
- search space
- benchmark problems
- optimal solution
- linear programming
- search algorithm
- markov decision process
- expected reward
- simulated annealing
- metaheuristic
- finite state