0
$\begingroup$

I am interested in knowing about algorithms for finding global minimum when noisy estimates of the function are available.

$\endgroup$
  • $\begingroup$ possible dup: cs.stackexchange.com/q/39546/755. What research have you done? What methods have you considered? Also take a look at cs.stackexchange.com/a/13343/755. $\endgroup$ – D.W. Sep 5 '15 at 4:37
  • $\begingroup$ @D.W.: I am aware of annealing kind of methods, but that gives convergence in probability rather than a.s. convergence. $\endgroup$ – RIchard Williams Sep 5 '15 at 6:23
  • $\begingroup$ That information ought to be in the question. Also, it suggests you have a requirement that is not stated (e.g., a.s. convergence). Second: the comment puzzles me. I had the impression that things like simulated annealing are heuristics, with almost no meaningful guarantees. Is this incorrect? Also, I'm puzzled by reliance on asymptotic notions of convergence: in practice we usually care more about whether it will give a good answer within our lifetime. Could you expand the question to elaborate more on the perspective you are approaching this from and how you will evaluate answers? $\endgroup$ – D.W. Sep 5 '15 at 15:38

Your Answer

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

Browse other questions tagged or ask your own question.