FMR-GA – A Cooperative Multi-agent Reinforcement Learning Algorithm Based on Gradient Ascent

2017 ◽  
pp. 840-848 ◽  
Author(s):  
Zhen Zhang ◽  
Dongqing Wang ◽  
Dongbin Zhao ◽  
Tingting Song
Keyword(s):  
Reinforcement Learning ◽  
Learning Algorithm ◽  
Gradient Ascent ◽  
Multi Agent ◽  
Reinforcement Learning Algorithm

Using Expectation-Maximization for Reinforcement Learning

1997 ◽  
Vol 9 (2) ◽  
pp. 271-278 ◽  
Author(s):  
Peter Dayan ◽  
Geoffrey E. Hinton
Keyword(s):  
Reinforcement Learning ◽  
Expectation Maximization ◽  
Learning Algorithm ◽  
Stochastic Gradient ◽  
Gradient Ascent ◽  
Relative Payoff ◽  
The Mean ◽  
Stochastic Gradient Ascent ◽  
Reinforcement Learning Algorithm

We discuss Hinton's (1989) relative payoff procedure (RPP), a static reinforcement learning algorithm whose foundation is not stochastic gradient ascent. We show circumstances under which applying the RPP is guaranteed to increase the mean return, even though it can make large changes in the values of the parameters. The proof is based on a mapping between the RPP and a form of the expectation-maximization procedure of Dempster, Laird, and Rubin (1977).

Sign in / Sign up

Export Citation Format

Share Document