3
$\begingroup$

A few days ago I created the tag . I did not discuss this tag on meta before creating it (although I mentioned creating this tag in the Tagging chatroom). Not so long after the tag creation a post in the Tag cleanup thread about this tag was made.

My reasons for creating this tag were:

  1. Although do not know anything about Kalman filter, judging simply by the fact, that I have heard it mentioned many times and also by the number of post here on MSE, it seems to be rather important topic.
  2. Many people kept using as a tag for such questions. However, as stated in the tag-info, this tag is supposed to be used for filters in the set-theoretic/order-theoretic sense. I hoped that creating a separate tag could help lower the number of mistagged questions.

As I am not familiar with Kalman filtering, I was not sure what to put into the tag-excerpt/tag-wiki, so I left them empty. (If someone can volunteer write something up, it would be great.)

My main intention was to solve the problem with incorrectly tagged question. This could be also solved by making a synonym of a more general umbrella tag. However, I am not sure whether there is an appropriate general tag for this. (I suppose that users asking about Kalman filtering would notice and use it instead of . If it is made a synonym of a more general tag, (kalman-filter) will be automatically changed to the master tag.)

Perhaps the most important question, which was already raised by Asaf, is: Is this tag useful? Should we keep it?

$\endgroup$
3
$\begingroup$

I don't think there should be a separate tag for every kind of filter that our colleagues at DSP think up. I suggest to capture all these.

(I also considered , , and , but these names would be specific to DSP and exclude applications to econometrics.)

$\endgroup$
  • $\begingroup$ At that site there also is kalman-filters tag. (Which, of course, does not mean that we need to have it too, since the other site is more specialized in this direction.) $\endgroup$ – Martin Sleziak Oct 6 '14 at 17:59
  • $\begingroup$ I like the suggestion for (data-filtering). I think that we might still keep (kalman-filter) and make it a synonym of the master tag (data-filtering). (It will increase the probability of adding the correct tag for questions about Kalman filter.) $\endgroup$ – Martin Sleziak Oct 7 '14 at 6:45
  • $\begingroup$ I don't think data filtering is an expression that anyone asking about Kalman filters would default to. A more general expression would be "stochastic filtering" - it's about updating estimates for stochastic processes/in stochastic calculus; but in my experience, it's really referred to as "filtering." Couldn't it be "filters (set theory)" and "filters (stochastic)", with tool tip? $\endgroup$ – gnometorule Oct 7 '14 at 16:44
  • $\begingroup$ @gnometorule Thanks for the input. Tags do not support spaces, and I don't think we have any with parentheses either. I like stochastic-filtering. $\endgroup$ – user147263 Oct 7 '14 at 16:47
  • 1
    $\begingroup$ @gnometorule: Note that the set theoretic filters are really just order theoretic filters, applied to $\subseteq$. And for many people that would really just be a topological issue, and not a set theoretical one, since one is likely to meet filters in a topology course. $\endgroup$ – Asaf Karagila Oct 8 '14 at 17:57
  • $\begingroup$ @Martin, CareBeer: do any of you watch the recent tags list? I'm a bit too lazy at the moment, but there has been a handful that seem to beg for removal. $\endgroup$ – Asaf Karagila Oct 11 '14 at 22:36
  • $\begingroup$ @AsafKaragila "euclidean" is bad; I retagged. (blaschke-products) is specific and unlikely to be misused; there are 50+ questions about Blaschke products on the site. I don't see it as harmful. There are others where I have no clue. $\endgroup$ – user147263 Oct 11 '14 at 23:08

You must log in to answer this question.

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