On the Use of Non-Stationary Policies for Stationary Infinite-Horizon Markov Decision Processes
Bruno ScherrerBoris LesnerPublished in: CoRR (2012)
Keyphrases
- non stationary
- infinite horizon
- markov decision processes
- optimal policy
- markov decision process
- finite horizon
- average cost
- dynamic programming
- state space
- finite state
- policy iteration
- decision problems
- stationary policies
- discount factor
- long run
- reinforcement learning
- control policies
- partially observable
- total reward
- decision processes
- reward function
- holding cost
- multistage
- single item
- average reward
- state dependent
- partially observable markov decision processes
- lost sales
- policy iteration algorithm
- action space
- markov decision problems
- reinforcement learning algorithms
- initial state
- expected reward
- sufficient conditions
- decision theoretic planning
- planning under uncertainty
- optimal control
- inventory level
- optimal solution
- decision making