I have a question about my Mathematics Stack Exchange post: What are the ways to analyze a random set of instances on a timeline?

Two weeks ago I asked this question to get help on the problem I have been working on for months. The question had zero answers so I put a bounty on it. A week had passed and I awarded the only answer with the bounty and accepted the answer. The answer was too vague to be real help for me but I wanted to reward the member's participation.

Today I revisited my question and to my surprise found that it was put on hold with the reason unclear what you're asking by Dominik, JonMark Perry, Shailesh, Jonas, user1551.

I did not ask my question out of curiousity or because I wanted attention. I am solving a real life problem. But since my question does not fall within the realm of conventional math class questions I am denied help altogether.

I have trade secrets to consider and I cannot go into much detail which might make my question sound broad. But why should my question be removed if a bunch of people cannot understand it?

  • 10
    $\begingroup$ Your question has not been removed, it has been put on hold. Every member of that "bunch of people" has earned the right, through contributions to this website, to vote to put a question on hold if he/she finds it unclear. If five qualified people find your question unclear, then it's up to you to clarify the question, if you want it reopened. $\endgroup$ – Gerry Myerson Dec 19 '16 at 7:02
  • 2
    $\begingroup$ I don't know what motivated the close voters, but IMHO the question is a bit vague. For example, you won't tell us what type of data you are looking at. Is it numerical? Is it a bunch of yes/no values? Is it a stream of characters? The available set of tools most likely depends heavily on the type of your data. If it is numerical you can begin by doing Fourier analysis on it, and catch nearly repeating structures as spikes in the Fourier transform. But, statisticians have written several books on analysing time series. $\endgroup$ – Jyrki Lahtonen Dec 19 '16 at 7:25
  • 1
    $\begingroup$ (cont'd) Which means that the question, as of now, might also qualify to be closed as too broad. Anyway, if the data consists of characters, then numerical methods lose a lot of their potency, because then you might need a way to measure the relative distance of two characters (by whatever metric you would decide they qualify for a pattern). If the data is just a sequence of bits, then you can study things like linear complexity, look for recurrence relations and such. $\endgroup$ – Jyrki Lahtonen Dec 19 '16 at 7:30
  • $\begingroup$ Having said all that I do realize that you may be looking for help exactly in how to mathematically model/measure whatever it is that you are studying. It is not unheard of that users here vote to close a question as unclear simply because they think they should understand it but they don't. But, I'm not sure that is the case here. Remember, the voters need to pick a reason for their close vote from a choice of very few alternatives. Therefore the message given by the system is just a best guess. $\endgroup$ – Jyrki Lahtonen Dec 19 '16 at 7:33
  • $\begingroup$ To get help to a specific problem you should describe the problem as specifically as you can. Talking only in generalities is likely to make it too broad. With a more detailed description some of us can also tell whether you may get better help at, say, CrossValidated. $\endgroup$ – Jyrki Lahtonen Dec 19 '16 at 7:35
  • $\begingroup$ @JyrkiLahtonen, I have trade secrets to consider, I cannot be specific. $\endgroup$ – user396522 Dec 19 '16 at 7:47
  • $\begingroup$ Retagging because support is for asking technical support in using the site software. $\endgroup$ – Jyrki Lahtonen Dec 19 '16 at 8:05
  • 7
    $\begingroup$ And what Gerry Myerson said. The question has not been removed. Re: Trade secrets. The more specific the question the more likely it is to get an answer (and be an acceptable question in the first place). Your call! Without a specific question you may not get any help. The community largely operates on the tradition of academic openness ... but whenever financial interests play a role it may be more difficult to get help at least without a share of the profits:-) That was a joke, don't even contemplate offering, such an offer would get deleted rather speedily. $\endgroup$ – Jyrki Lahtonen Dec 19 '16 at 8:15
  • $\begingroup$ Asking a meta question attracted a wave of downvotes without any comments - what a helpful and friendly community this is. $\endgroup$ – user396522 Dec 19 '16 at 14:27
  • 7
    $\begingroup$ Hi, one of the "bunch of people" here. When I cast the vote, I did --- and still do --- think that the question is unclear. E.g. are the time points discrete or arbitrary real values? By an "interval", did you mean the interval between two consecutive instances or between any two instances? (The first list item in your question seemed to suggest the former, but the $O(n^2)$ complexity seemed to suggest the latter.) I voted to put the question on hold, so that you could clarify it and people wouldn't give answers that do not address the question you have in mind. $\endgroup$ – user1551 Dec 19 '16 at 16:58
  • 6
    $\begingroup$ Often, one should comment on what are expected to be clarified, but seeing that you had awarded a bounty to an answer, I thought you'd got what you needed and so I didn't comment. In hindsight, that was probably a mistake. By the way, I didn't down-voted any of your questions, but on MSE, a down-vote to a meta question simply means disagreement rather than dislike. E.g. one may down-vote your meta question because he/she disagrees that putting a question on hold means removal. Even top users may get a lot of down-votes here. Please don't get emotional on it. $\endgroup$ – user1551 Dec 19 '16 at 17:03
  • 11
    $\begingroup$ If there are trade secrets to consider, hire a mathematician. This community is helpful and friendly to those who don't come here looking for ways to make money off us. $\endgroup$ – Gerry Myerson Dec 20 '16 at 3:15
  • 1
    $\begingroup$ @GerryMyerson, I have an employer and business ethics. I was trying to get help but instead I am being accused by you of trying to exploit members. That is indecent and offensive. $\endgroup$ – user396522 Dec 20 '16 at 6:48
  • 9
    $\begingroup$ You've been here two weeks, Robert, and think you know better how the site should work than people who have contributed to it for years. That is indecent and offensive. $\endgroup$ – Gerry Myerson Dec 21 '16 at 4:15

When a user does not know how to ask a question in mathematical symbols, I generally prefer that they get help in person. For a student, this would be a tutor, perhaps paid.

In business, this would be a consultant, and cost more.

The trouble with questions that are too broad is that there is little reason to think that a person who cannot ask a focused question will be able to correctly implement the answers. Better for everyone if you are able to hire a mathematics/statistics graduate student or similar, get your question refined, and get to the point where you are doing the whole thing correctly.

I remember a doctor near Denver who asked here for significant statistics help, and I was in favor of the project. the University of Denver has explicit consulting fees for the Statistics Department, as this happens with some frequency. Professors are expensive, graduate students less.

Anyway, you say you have a time series, and trade secrets. Not going to work well on a website.


You must log in to answer this question.