Reinforcement Learning under a Multi-agent Predictive State Representation Model: Method and Theory.
Zhi ZhangZhuoran YangHan LiuPratap TokekarFurong HuangPublished in: ICLR (2022)
Keyphrases
- reinforcement learning
- mathematical model
- multi agent
- evaluation method
- objective function
- statistical model
- predictive model
- probabilistic model
- graph representation
- modeling method
- theoretical analysis
- prior knowledge
- control policy
- prediction model
- cost function
- model free
- bp neural network
- spatial structure
- mathematical framework
- test data
- state space
- classification method
- transition model
- optimization method
- sensitivity analysis
- linear regression
- reconstruction method
- linear model
- em algorithm
- matrix representation
- input data
- computational model
- tree structure
- prior information
- machine learning
- energy function
- theoretical framework
- representation scheme
- support vector machine
- dynamic programming
- pairwise
- adaptation process