CS294-112 深度强化学习 秋季学期（伯克利）NO.5 Actor-critic introduction

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

in most AC algorithms, we actually just fit value function. less common to fit Q function as well.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

batch：off line， monte carlo。online： bootstrap，TD

 

 

 

in fast emulator，use the left one

 

 

 

 

 

 

 

 

 

 

 

 

 

this strategy works well in the beginnning of training