An Evolutionary Approach to Find Optimal Policies with an Agent-Based Simulation.
Nicolas De BufalaJean-Daniel KantPublished in: AAMAS (2019)
Keyphrases
- optimal policy
- markov decision processes
- decision problems
- reinforcement learning
- state space
- finite horizon
- infinite horizon
- multistage
- dynamic programming
- finite state
- long run
- average reward
- state dependent
- markov decision process
- bayesian reinforcement learning
- dynamic programming algorithms
- sufficient conditions
- markov decision problems
- policy iteration
- average cost
- initial state
- machine learning
- serial inventory systems
- reward function
- model free
- optimal solution
- learning algorithm