UPDeT: Universal Multi-agent RL via Policy Decoupling with Transformers.
Siyi HuFengda ZhuXiaojun ChangXiaodan LiangPublished in: ICLR (2021)
Keyphrases
- multi agent
- reinforcement learning
- optimal policy
- learning agents
- rl algorithms
- markov decision process
- action selection
- multi agent systems
- partially observable markov decision processes
- policy search
- sequential decision making
- control policy
- cooperative
- single agent
- input output
- policy gradient
- control policies
- state space
- intelligent agents
- function approximators
- policy evaluation
- action space
- markov decision processes
- approximate dynamic programming
- machine learning
- state and action spaces
- partially observable
- reinforcement learning algorithms
- autonomous agents
- model free
- coalition formation
- reinforcement learning problems
- average reward
- state action
- markov decision problems
- actor critic
- multi agent environments
- partially observable domains
- multiagent reinforcement learning
- heterogeneous agents
- continuous state spaces
- policy iteration
- complex environments
- multiagent systems
- dynamic programming