Action-depedent Control Variates for Policy Optimization via Stein's Identity

  title={Action-depedent Control Variates for Policy Optimization via Stein's Identity},
  author={Hao Liu and Yihao Feng and Yi Mao and Dengyong Zhou and Jian Peng and Qiang Liu},
Policy gradient methods have achieved remarkable successes in solving challenging reinforcement learning problems. However, it still often suffers from the large variance issue on policy gradient estimation, which leads to poor sample efficiency during training. In this work, we propose a control variate method to effectively reduce variance for policy gradient methods. Motivated by the Stein’s identity, our method extends the previous control variate methods used in REINFORCE and advantage… CONTINUE READING


Publications referenced by this paper.
Showing 1-10 of 39 references

Approximate computation of expectations

  • Charles Stein
  • Lecture Notes-Monograph Series,
  • 1986
Highly Influential
4 Excerpts

Similar Papers

Loading similar papers…