4
$\begingroup$

I just noticed that we have tag with 18 questions - all about Bayesian updating. Do we really need a separate tag? Maybe we should merge it with ?

$\endgroup$
7
  • 1
    $\begingroup$ My first thought is yes. I wonder if there is value in having a tag to pick out questions that are specific to updating, as opposed to other aspects of Bayesianism, but updating seems pretty central to B. I'm not sure if B is terribly meaningful w/o U; if not, I see no reason to have a separate tag. $\endgroup$ Sep 1 '16 at 15:20
  • 6
    $\begingroup$ My first thought on seeing "updating" was Choleski updating, then Kalman filter updating ... Bayesian updating didn't occur to me. I'm not quite sure merging it with Bayesian makes sense to me, though, since "Bayesian updating" is more specific -- it conveys infomation. Can we rename a tag? I'd prefer not to have such an ambiguous tag as "updating" without qualifying it. $\endgroup$
    – Glen_b
    Sep 1 '16 at 15:50
  • 2
    $\begingroup$ I agree with Glen_b that [updating] is so ambiguous that we should not make it a synonym of anything; the tag has to be removed altogether, or renamed. I'd agree with @gung that simply replacing it with [bayesian] is fine. Renaming it into [bayesian-updating] is another alternative, but then one should look at what other bayesian threads deserve such a tag. This search suggests that quite a bit. $\endgroup$
    – amoeba
    Sep 1 '16 at 18:50
  • 1
    $\begingroup$ @Glen_b I recall that tag can be renamed via the tag merge interface. If not, then in any case one can merge A into B and then remove the synonym mapping, thus effectively renaming A into B without leaving any traces. $\endgroup$
    – amoeba
    Sep 1 '16 at 18:51
  • 3
    $\begingroup$ With only 18 questions, we could also just remove them all & let the tag die. $\endgroup$ Sep 1 '16 at 18:55
  • $\begingroup$ @amoeba thanks, yes. That's what I get for posting when I should be asleep $\endgroup$
    – Glen_b
    Sep 1 '16 at 23:17
  • 1
    $\begingroup$ I would be in favor of eliminating the updating tag. Note that there in addition to the bayesian tag, there is also a posterior tag, which is perhaps also relevant as a mapping-target? (On Glen_b's comment, I agree there are other meanings for updating, e.g. for matrix factorizations. However I believe "Kalman filter updating" is just a particular instance of bayesian updating.) $\endgroup$
    – GeoMatt22
    Sep 6 '16 at 0:06
8
$\begingroup$

This answer is based on the discussion in the comments above.

The tag is so ambiguous that we should not make it a synonym of anything; the tag has to be removed altogether. With only 18 threads, the easiest is to manually remove it and let the tag die.

Renaming it into [bayesian-updating] is another possible alternative, but I do not see that this would be such a useful tag, so I'd suggest we simply eliminate it.


Update: I have eliminated the tag.

$\endgroup$

You must log in to answer this question.

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