Train Small, Deploy Big: Do Relative World Views Permit Swarm-Safety During Policy Transplantation for Multi-Agent Reinforcement Learning Problems?
Bradley FraserGiuseppe LauritoPublished in: Australasian Conference on Artificial Intelligence (2020)
Keyphrases
- reinforcement learning problems
- multi agent
- reinforcement learning algorithms
- reinforcement learning
- reinforcement learning methods
- natural actor critic
- function approximation
- function approximators
- policy iteration
- particle swarm optimization
- cooperative
- markov decision problems
- model free
- multi agent systems
- optimal policy
- markov decision processes
- dynamic programming
- temporal difference
- action space
- machine learning
- single agent
- evolutionary algorithm
- infinite horizon
- monte carlo
- supervised learning