Hot answers tagged

26

I'm in favor of the merger. Any questions which are specifically about deep neural networks, in contrast to neural networks in general, can be described as such in the text. One motivation for the distinction in the literature is that training a network with more layers can be harder than training a network with fewer, among other similar types of ...


23

It is on-topic. Because every comment so far could be an interesting answer to this question I am pasting them here: Nick Cox: The tag or term data-preprocessing isn't universal across statistical science (minimally, it's not a term I meet around the place or in literature), but whether and how to deal with outliers, work on transformed scales, etc. seem ...


18

Statistical questions are unlike mathematical questions. On the math site, a question will quickly be closed if it is not immediately clear what it means and what it is asking. All other questions on that site might as well be textbook questions--there is really no basis for distinction, not even the level of sophistication. Our community has learned to ...


17

This is IMO totally reasonable. In fact, I've seen other sites organize tag clean-ups to try to get the community involved in similar behavior. About the only thing to be concerned with is mass retagging without community approval. Users are just expected to use their best judgement about the boundaries between mass retagging and cleaning up a few questions....


17

Technical statistical usage sometimes departs from colloquial English usage. (Examples abound, such as the transitive use of "to learn" in the ML community and the use of singular nouns among statisticians to refer to entire families of distributions. And that's not even touching on the severely technical meanings of "confidence," "error," "fit," "...


16

Thanks for bringing our attention to that! I have removed this tag from the three threads. Obviously, we don't need such a tag on Cross Validated as it targets a particular UI specific to R (even if this is probably the most useful and least intrusive :-).


16

The master tag should be covariance-matrix.


16

After some discussion in the comments above, the emerging consensus seems to be: Map gradient-boosted-trees and gbm to boosting as synonyms. Leave adaboost and xgboost alone. This plan looks good to me. Upvote if agree, downvote (and explain in the comments!) if disagree. Update (Nov 5): This has been implemented.


15

Make deep-rl a synonym of reinforcement-learning. If users want to emphasize that their network is "deep," they have the option to add deep-learning.


14

One reason students-t can potentially keep reappearing is that it starts with "Student". We don't have any other tags starting with this word. So instead of blacklisting this tag, I suggest to create students-t-test as a synonym of t-test and students-t-distribution as a synonym of t-distribution. Then whoever starts typing "student..." in the tag field ...


14

Based on number of posts: Time of day: Looks like 3:00-7:00 (all times are UTC) is half as busy as 14:00-18:00 day of week: Saturday is the least busy day, followed by Sunday combined: 6:00-7:00 Saturday is just over 20% as busy as 15:00-16:00 Wednesday via this query -- http://data.stackexchange.com/stats/query/51030/posts-by-day-of-week-and-hour-of-...


13

I am a little hesitant to write this "answer", but let me explore a question of a slightly broader nature. Tagging. In my view, a tag should be indicative of the subject matter and we should avoid so-called "meta tags" (e.g., homework) insofar as possible and reasonable. I've seen the resurgence of a couple that seem borderline cases of this recently. I ...


13

I'm going to dispute some of the "facts" as presented in the SO version of this argument, which I think either don't apply here, or are just a little bit wrong. Generally speaking, I favor retaining the homework tag. "homework is a meta-tag". I don't necessarily think meta-tags are bad. It helps put the question in context - this is homework, or an ...


13

This is the output from a social network analysis. (I don't know much about SNA, but we have a tag for it, social-network.) Specifically, this is an undirected graph where the sizes of the nodes are proportional to the number of questions with that tag (actually, it's probably the log or square root, etc., of the number [no it isn't, note the comment from @...


13

Create the tag, don't mass-retag - retag slowly, one question at a time, not too many all at once, so we don't bump-flood the landing page.


13

There are only 14 threads tagged with science. I looked through them and I don't see anything in common among these threads. It appears to me that this tag (that, by the way, exists since 2010) is being applied without any coherent system and hence does not serve any useful purpose. I would suggest getting rid of this tag. Update (Jan 23): I have ...


13

I don't see why not. You make good points.


12

A risk of adding package tags would be 1) overlapping nomenclature 2) frequent disagreement on tags and 3) difficulty consulting relevant questions from other tagged questions with answers. For instance, a user working with the survey package would probably often mistakenly use the survey tag instead of the survey-package tag. If a user tags their question ...


12

Thanks. I merged visualisation and data-visualisation with data-visualization.


12

Yes it would. As it happens, the [combining-p-values] tag did not exist until around one week ago, when I created it and single-handedly tagged around 10 questions with it. There was some activity in several questions about Fisher's method and it prompted me to create this tag; I felt that [meta-analysis] is too general. I was not sure what is the best name ...


12

This seems fine to me. Here are a couple of points: When you create the tag, be sure to create a tag wiki excerpt that states what it stands for, and provides some guidance on how to use the tag. You may want to create a full tag wiki with more in-depth information as well. Don't add the tag to all the relevant threads at once. Do a few at a time ...


11

While multilevel and multilevel-analysis are probably easily mergeable, some of the others are genuine terminology differences. "Random effects" to economists is the GLS estimator of the random intercept model in the panel-data models. The "fixed effect" is a conditional estimator that only uses the within-cluster variability. I am sure that whatever is ...


11

Organizing tags is a delicate balancing act between being too comprehensive and too specific. It is no surprise, then, that discussions about tags occur frequently. I am grateful that our community is constantly thinking about them and proposing improvements. Tags are primarily intended to help people find information. Secondarily, they organize threads ...


11

No, we do not need a distinct rstudio tag because rstudio is simply a convenient interface to R, not a separate language with its own peculiarities. Making rstudio a synonym of r appears to solve any residual need to consolidate R-related tags.


11

I've argued the other way in the thread cited, both by commenting against this point of view and by answering the question. But we can agree that questions such as these do seem to straddle a border. The question is how much of the question is on each side. My position is pragmatic. If a usage question is only in small part about meanings or senses that ...


11

Like @Scortchi and @whuber, I prefer shorter excerpts. (Besides aesthetic preference, I suspect they are more likely to be read.) I think excerpts should mostly proffer tag usage guidance, and that more substantive and detailed information is better placed in the full wiki. That said, what is "shorter" exactly? To some degree "shortness" is in the eye ...


11

Yes, we need both. I think forecasting is more specifically about time-series data -- "What happens next?" Prediction the task of deploying a model to assess unseen data in some way. With prediction, there need not be any time-series component at all; you might just be interested in how well a model does against a holdout set of observations. Support of ...


11

Usage of "weights": regression coefficients Regression coefficients are referred to as "regression weights" in some literatures. I personally find this confusing, and would like to steer the site users to the more traditional statistical literature that talks about coefficients. Upvote: questions that talk about regression coefficients as weights should ...


11

Yes, mapping the function to the packages is a good idea. Maybe we should make all of them as synonyms to [lme4-nlme] ? This is the simplest choice and anyone trying to use one of the listed tags would not have problems with finding appropriate one since they all will be synonyms to the more general tag. Moreover, those functions are used in quite similar ...


11

The learning tag is in the process of being removed, see the answer in Understanding the use of the [education] tag. I wrote wiki excerpts to q-learning and transfer-learning. The former is IMHO a valid & useful tag, about the latter I am not sure as it only has 5 threads, but the concept seems clear enough so I'd say let it be. I removed [...


Only top voted, non community-wiki answers of a minimum length are eligible