Combining backpropagation with equilibrium propagation to improve an actor-critic reinforcement learning framework

HIGHLIGHTS

  • who: Yoshimasa Kubo and Artur Luczak from the Pukyong National University, Japan have published the article: Combining backpropagation with Equilibrium Propagation to improve an Actor-Critic reinforcement learning framework, in the Journal: (JOURNAL)
  • what: The authors explore an Actor-Critic architecture trained by both BP and by braininspired modification of EP proposed by Luczak et_al . This study provides two contributions: Actor-Critic is a two-part architecture for reinforcement learning. The authors demonstrate that the architecture can solve several control tasks, and that its learned behaviors are more consistently rewarding than behaviors learned using BP . . .

     

    Logo ScioWire Beta black

    If you want to have access to all the content you need to log in!

    Thanks :)

    If you don't have an account, you can create one here.

     

Scroll to Top

Add A Knowledge Base Question !

+ = Verify Human or Spambot ?