131

I'm currently working on many machine learning and computational algorithms, such as Singular Value Decomposition, Support Vector Machines, and others.

I'd like to ask questions about these topics, but I don't know which Stack Exchange website to use to post my questions.

How do I tell? What criteria do I use?

4
  • 3
  • 4
    Or should you post in Artificial Intelligence?
    – Lii
    Jun 2, 2020 at 6:45
  • I think these resources have setup a ML categorizing algorithm that filters out ML questions because sometimes the reason for taking questions down doesn't even make any sense in the context of the question. It makes me think that the decision is not made by humans. I am struggling with this for months now.
    – MsTais
    Dec 1, 2021 at 22:10
  • In ComputerScinece the activity is very low as compared to some other portals. My question was not taken down there though... maybe for the same reason X)))
    – MsTais
    Dec 1, 2021 at 22:23

7 Answers 7

62

I've ended up here after asking this on MetaStackOverflow, as the question of where to send "pure Machine Learning" offtopics remains unclear (to me). My initial guess was that the place to put this stuff is CrossValidated (stats), but here is my analysis on the matter:

There are several possibilities for asking ML questions right now (ordered by site traffic):

  1. CrossValidated: enter image description here

    a) It includes machine learning in its main topics:

    Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization.

    b) machine-learning is the 4th most popular tag in the site (3359 tagged questions).

  2. ComputerScience (beta) enter image description here

    a) It mentions "Machine Learning" in its help page, but not in its main page:

    Computer Science Stack Exchange is a question and answer site for students, researchers and practitioners of computer science.

    b) machine-learning is the 22th most popular tag. The site seems to be more about algorithms.

  3. Computational Science is not related to Machine Learning according to its Help page.

  4. Theoretical Computer Science enter image description here

    a) According to its Help Center,

    TCS covers a wide variety of topics including algorithms, data structures, computational complexity, parallel and distributed computation, probabilistic computation, quantum computation, automata theory, information theory, cryptography, program semantics and verification, machine learning, computational biology, computational economics, computational geometry, and computational number theory and algebra.

    Work in this field is often distinguished by its emphasis on mathematical technique and rigor.

    On the other hand, it also says:

    For questions other than research-level questions in TCS, you may want to consider the following places to ask:

    General Artificial Intelligence — Meta Optimize

    Statistics and Data Mining — Cross Validated ...

    What I understand of all this is (and the word "Theoretical" seems like a hint) is that TCS is the right place for theoretical machine learning questions, leaving practical questions on the side.

    b) machine-learning is the 24th most popular tag in the site.

  5. Data Science enter image description here

    This claims to be a place for "machine learning professionals", and machine-learning is the most popular tag! (263 questions).


Given all this information, I see there are 2 different options:

a) If the question is practical ("How do I split my 5000 sample training set to improve my SVM performance"), I see two possibilities: CrossValidated and Data Science. Since the final goal is to get help, I'd rather go for CrossValidated as it has a considerable traffic and machine learning seems like a relevant topic in there.

b) If the question is about machine learning theory (for instance, understanding how a neural network or SVM works), there are two natural options: Computer Science or Theoretical Computer Science. If the question is research-level and theoretical (e.g., questions about PAC learning, provable performance guarantees), ask on Theoretical Computer Science. If the question is not research-level, ask on Computer Science.

7
  • 1
    I would add a final option (c) reminding that programming-related machine-learning questions do belong on Stack Overflow. After all we already have more that 6300 ML questions on SO...
    – Amro
    Apr 22, 2015 at 12:47
  • That's true, but I'm talking about **pure**(not programming-related) ML questions...
    – Яois
    Apr 22, 2015 at 15:33
  • 1
    There's a problem with your bottom-line advice a). TCS is only for research-level questions. Therefore, the examples you give are not suitable for TCS. I've edited your final recommendation to recommend either TCS or CS for theory questions, depending on whether the question is research-level.
    – D.W.
    Apr 22, 2015 at 20:06
  • 2
    Other feedback: It sounds like you've concluded that Stats.SE is not a great place for questions about machine learning theory. Do you want to elaborate on the basis for that conclusion? In my experience Stats.SE often seems open to theoretical/conceptual questions as well (though they might come from a slightly different perspective than computer scientists).
    – D.W.
    Apr 22, 2015 at 20:48
  • 8
    The fact that this question took so long to answer (and that I'm still confused after reading it) tells me that something is wrong... Mar 5, 2018 at 11:39
  • 1
    Wouldn't it be opportune to open a dedicated 'machine learning' SE website in order to centralize question about frameworks, theoretical and practical aspects? I'm freely asking...
    – s.k
    Apr 15, 2019 at 9:52
  • CrossValidated just took my question down. In fact, stackoverflow.com even banned me for posting ML questions. They think my minimal codes were not minimal enough...
    – MsTais
    Dec 1, 2021 at 21:53
22

Should I use Computer Science one? Should I use Theoretical Computer Science one? Should I use Computational Science one? Should I use Statistical Analysis one?

Machine learning should be on-topic for either Computer Science or Cross Validated. If your questions are at least graduate study level and theoretical (as explained in their scope), then they should be acceptable at Theoretical Computer Science as well. Do any of those sites have questions similar to the ones you want to ask that are currently getting answered? If so, then that's the one I'd pick.

0
12

Questions about use and selection of machine learning algorithms, or about machine learning generally, are appropriate for DataScience.SE, where machine-learning has been the most popular tag in the first month of private and public beta.

However, questions about implementing or design of algorithms are probably more appropriate for one of the computer science sites that @BilltheLizard mentioned. See also: Are questions about algorithm implementation on-topic?

4

Research level questions should go to Theoretical CS, anything lower should go to CS. If it's a question about a statistical method, either within the model or analyzing output, and it doesn't require any CS knowledge, I'd think Cross-validated (stats) would be best.

There have been a couple area 51 proposals for AI and ML that have actually made it to beta, but both were shut down after private, I believe due to a lack of experts and expert-level questions.

3

Note that there are also several Area51 sites/proposals about Artificial Intelligence or Machine Learning. For example here.

1
  • Artificial Intelligence SE has been live for some time now.
    – desertnaut
    Nov 23, 2020 at 23:36
2

One way to select the site is to think what main tag you will use and then check popularity of that tag in each of the sites.

For example, if you want to ask about feature engineering, your results will be:

  1. Computer Science

'feature' tags in computer science

  1. Theoretical Computer Science

'feature' tags in theoretical computer science

  1. Computational Science

'feature' tags in computational science

  1. CrossValidated

'feature' tags in cross validated

  1. Data Science

'feature' tags in data science

  1. StackOverflow

'feature' tags in stackoverflow

Based on that, I'd say that feature engineering questions should go to CrossValidated or Data Science.

0

DataScience and Artificial Intelligence will do.

You must log in to answer this question.

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