0
$\begingroup$

Yesterday I posted this question on Mathematics Stack Exchange. Except for the one upvote it didn't receive any attention. What is the problem with my question? Is it too vague? Should I better include an example? I am hesitating whether I should have posted the question on MathOverflow instead of this SE. Should I change the location of the question?

| |
$\endgroup$
  • $\begingroup$ (1) In a site with that much traffic, getting one upvote is something to begin with. (2) Why do you think this question should be on MathOverflow? $\endgroup$ – Asaf Karagila Jan 23 '16 at 15:32
  • $\begingroup$ Because it is not a typical mathematics question that can be answered with a proof or a calculation. $\endgroup$ – Ruben Jan 23 '16 at 15:40
  • $\begingroup$ And you think that all questions on this websites are just questions that can be answered by calculations or proofs? Or that the questions on MathOverflow are different somehow? $\endgroup$ – Asaf Karagila Jan 23 '16 at 15:41
  • $\begingroup$ By the tone of your question, I guess the answer is no... $\endgroup$ – Ruben Jan 23 '16 at 15:43
  • $\begingroup$ Last but not least, this is not a good reason to delete that question. $\endgroup$ – user26857 Jan 24 '16 at 9:45
  • $\begingroup$ But I posted the question on another stackexchange? $\endgroup$ – Ruben Jan 24 '16 at 12:23
3
$\begingroup$

The mere fact that a question does not get much attention does not in itself mean the question is lacking. The subject you ask about is not one that gets that much attention in general on this site.

However, I feel your question is also quite broad and possibly a bit hard to parse. The introduction of $t$ is not quite transparent, but as it is not my field and I still can sort of get the point it should be fine. But it is still quite broad a question.

To ask this question on MO does not strike me as a good idea, for one thing as this type of mathematics is not that well-represented on MO either.

If I had this type of question I would ask it on the site for Computational Science. The point that the question is a bit broad rests though.

If you want some specific advice: try to focus the question a bit more, and ask it on that site.

| |
$\endgroup$
  • 1
    $\begingroup$ That Computational Science website looks indeed interesting. I will post it there. About my question, I had hoped that there would be a general answer, but it's clear that I should be more specific. $\endgroup$ – Ruben Jan 23 '16 at 15:51
  • $\begingroup$ I am glad you found the information helpful. I think there is a very general answer, in that if you have one variable and then/thus consider say a subset of $[0,1]$ and you descretise to have to have a distance of $.01$ than you have about $100$ points (technically $101$). If you . have two, you'd look at $[0,1]^2$ and then you'd need $100^2=10^4$. points for about the same resolution. With $5$ vqariables youd be at $10^10$ and so on. The growth in the dimension is exponential. See en.wikipedia.org/wiki/Curse_of_dimensionality But with a more focused q you'll get better answers. $\endgroup$ – quid Jan 23 '16 at 16:19
2
$\begingroup$

I think the question is rather vague and would be improved by including an example. Right now, the answer would appear to be "Well, it depends on what $f$, $t$ and $g$ are" - which isn't a very compelling answer. I'd bet it also depends on the method you're using for numerical optimization. In any case, being more specific would make the question substantially more answerable.

Also, it's worth noting that not receiving much attention isn't necessarily because there's anything wrong with the question. Sometimes that's just how it is.

(And, I'm not a regular on MO, but it doesn't look like a question about research level mathematics; it seems somewhat elementary, just vague)

| |
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .