Finite-Horizon and Infinite-Horizon Markov Decision Processes with Trapezoidal Fuzzy Discounted Rewards.
Karla Carrero-VeraHugo Cruz-SuárezRaúl Montes-de-OcaPublished in: ICORES (Selected Papers) (2021)
Keyphrases
- markov decision processes
- finite horizon
- infinite horizon
- optimal policy
- markov decision process
- state space
- stochastic demand
- finite state
- single item
- production planning
- reinforcement learning
- inventory control
- average cost
- dynamic programming
- average reward
- policy iteration
- partially observable
- single product
- action space
- reward function
- long run
- control policies
- total reward
- inventory policy
- discounted reward
- lost sales
- decision problems
- multistage
- periodic review