Towards Designing Optimal Reward Functions in Multi-Agent Reinforcement Learning Problems.
Ricardo GrunitzkiBruno Castro da SilvaAna L. C. BazzanPublished in: IJCNN (2018)
Keyphrases
- reward function
- multi agent
- reinforcement learning algorithms
- reinforcement learning problems
- reinforcement learning
- multiple agents
- state space
- markov decision processes
- reinforcement learning methods
- function approximation
- state variables
- dynamic programming
- optimal solution
- optimal policy
- partially observable
- average reward
- policy search