Improving adaptability to new environments and removing catastrophic forgetting in Reinforcement Learning by using an eco-system of agents.
Olivier MoulinVincent François-LavetPaul W. G. ElbersMark HoogendoornPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- multi agent environments
- multi agent
- multi agent systems
- learning capabilities
- dynamic environments
- learning agents
- autonomous agents
- learning agent
- single agent
- action selection
- multi agent reinforcement learning
- multiagent systems
- multiple agents
- complex environments
- cooperative
- intelligent agents
- highly dynamic
- multiagent learning
- reinforcement learning agents
- open systems
- artificial agents
- interacting agents
- agent receives
- incremental learning
- machine learning
- robocup soccer
- reinforcement learning algorithms
- dynamic programming
- function approximation
- state space
- software agents
- agent behavior
- mobile agents
- mobile robot
- markov decision processes
- multiagent reinforcement learning
- partial observability
- agent systems
- temporal difference
- decision theoretic
- model free
- learned knowledge
- reinforcement learning methods
- stochastic games
- optimal policy
- evolutionary learning
- function approximators
- trust model
- autonomous robots
- decision making