CtRL-Sim: Reactive and Controllable Driving Agents with Offline Reinforcement Learning.
Luke RoweRoger GirgisAnthony GosselinBruno CarrezFlorian GolemoFelix HeideLiam PaullChristopher PalPublished in: CoRR (2024)
Keyphrases
- reinforcement learning
- reactive agents
- multi agent
- agent architecture
- learning agents
- multi agent systems
- action selection
- intelligent agents
- multi agent reinforcement learning
- autonomous agents
- agent behavior
- agent receives
- multiagent systems
- multi agent environments
- multiagent learning
- learning agent
- software agents
- single agent
- reinforcement learning agents
- multiple agents
- coalition formation
- real time
- partial observability
- complex environments
- robocup soccer
- function approximation
- dynamic environments
- learning algorithm
- cooperative
- state space
- reactive systems
- mobile agents
- partially observable
- resource allocation
- multi agent learning
- dynamic programming
- reinforcement learning algorithms
- agent learns
- agent technology
- markov decision processes
- trust model
- agent systems
- incomplete information
- evolutionary learning
- multiagent reinforcement learning
- decision making
- temporal difference