Adaptive action supervision in reinforcement learning from real-world multi-agent demonstrations.
Keisuke FujiiKazushi TsutsuiAtom ScottHiroshi NakaharaNaoya TakeishiYoshinobu KawaharaPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- multi agent
- real world
- action selection
- wide range
- adaptive control
- action space
- state space
- model free
- function approximation
- traffic signal control
- reward shaping
- reinforcement learning algorithms
- data mining
- learning agents
- intelligent agents
- machine learning
- multi agent environments
- state action
- cooperative
- data sets
- markov decision processes
- multi agent systems
- learning algorithm
- multi agent reinforcement learning
- reinforcement learning agents
- optimal policy
- fitted q iteration
- case study
- initial state
- synthetic data
- reinforcement learning methods
- transition model
- autonomous agents
- extensive form games