I wonder if someone could shed some light on this. I'm curious if stochastic parallel-gradient descent and simultaneous perturbation stochastic approximation refer to the same optimization techniques.

  • 1
    $\begingroup$ They do appear to be based on basically the same idea, developed and named by different authors. Of course, the devil is in the details, and specific detailed implementations might differ in significant ways, regardless of what the algorithm is called.. $\endgroup$ Sep 22 at 22:13
  • $\begingroup$ @MarkL.Stone, thanks. I read a few paper on this but I'm not clear the fine difference $\endgroup$
    – Young Wang
    Sep 23 at 13:46


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy