Breaking the Curse of Many Agents: Provable Mean Embedding Q-Iteration for Mean-Field Reinforcement Learning.
Lingxiao WangZhuoran YangZhaoran WangPublished in: ICML (2020)
Keyphrases
- reinforcement learning
- multi agent
- learning agents
- multi agent systems
- action selection
- multi agent environments
- intelligent agents
- agent receives
- autonomous agents
- function approximation
- multiagent systems
- multi agent reinforcement learning
- multiagent learning
- learning agent
- single agent
- learning capabilities
- multiagent reinforcement learning
- high dimensional
- dynamic environments
- cooperative
- robocup soccer
- decision making
- artificial agents
- reinforcement learning agents
- multiple agents
- software agents
- agent model
- resource allocation
- complex environments
- temporal difference
- mobile agents
- machine learning
- partial observability
- coalition formation
- data hiding
- belief networks
- agent architecture
- linear complexity
- closed form
- optimal policy
- high dimensional data
- statistical mechanics
- markov random field
- dimensionality reduction
- np hard
- bayesian networks
- model free