I have been reading some medium articles and these three methods pop up a lot. I am wondering what the differences between these are, what are the advantages of one over the other, etc. Also from my understanding, actor-only method is synonymous to policy gradient methods, critic-only method is synonymous to value-based methods, and actor-critic methods use policy gradient methods for the actor part and value-based methods for the critic part. Am I correct? Any help would be greatly appreciated. Finally, what should one keep in mind when considering one or the other for their projects?

  • $\begingroup$ You're asking many questions here. I think they can barely be in the same post, provided that you edit your post to emphasize what your main question is and the other questions are just sub-questions that would also help to answer the main one. $\endgroup$
    – nbro
    Commented Jul 2, 2022 at 23:24
  • $\begingroup$ Hi, thanks for the reply. My main question would be what the difference is between these three: actor-only, critic-only, and actor-critic methods. $\endgroup$ Commented Jul 4, 2022 at 10:21


You must log in to answer this question.