Which site is better to ask question about Machine Learning, Mathematics meta or cross validated?
-
7$\begingroup$ Mathematics meta would definitely not be the place -- meta is a site for asking questions about Math.SE. $\endgroup$ – Nick Peterson Mar 19 '14 at 12:42
-
$\begingroup$ Machine learning means: machines used in teaching humans, right? or does it mean machines who learn things.... $\endgroup$ – GEdgar Mar 19 '14 at 12:58
-
4$\begingroup$ @GEdgar The latter, according to the tag wiki. $\endgroup$ – user127096 Mar 19 '14 at 14:07
-
$\begingroup$ Related on Meta Stack Exchange: Where to ask basic questions about machine learning? and Which StackExchange website for Machine Learning and computational algorithms?. $\endgroup$ – TooTone Apr 22 '14 at 19:31
-
-
1$\begingroup$ @Pinocchio Click tag name, then "top users". $\endgroup$ – user127096 Mar 19 '14 at 0:24
Recently, I've been using both sites quite heavily. The answer is it depends on what you are looking for. There are some very good mathematicians on both sites, but obviously a lot more of them on maths.se. Where stats.se really scores in my opinion, is when you have a question that is not just mathematical but also involves methodology (the clue is in the name: "cross-validation :), and/or interpretation of results.
In summary, if your questions can be tightly defined in terms of the mathematics, go to maths.se. If it also involves how to perform the analysis, issues related to what features of the data to choose, or how to interpret results, etc, go for stats.se -- and if there's mathematics involved the guys on stats.se can handle it. They're both great sites, and it's worth having a look at questions and answers on each one to get a flavour of how each works. Whatever you do, don't cross-post.
There is a similar question on meta.stats.se.
So using cheap effective diet pills' heuristic, you should post on StackOverflow:
In addition to Mathematics, Cross Validated, and Stack Overflow there is a data science proposal currently in the commit phase that includes machine learning.
-
1$\begingroup$ Also twice an AI SE got created and deleted (and once for ML AFAIR). $\endgroup$ – Franck Dernoncourt Apr 1 '14 at 2:01