« first day (2185 days earlier)      last day (2823 days later) » 

Phrancis
Phrancis
1:33 AM
I removed those stdiv and var columns, and instead created some delta columns which may be more useful overall
 
 
6 hours later…
mark999
mark999
7:58 AM
Hope your interview went well @MatthewDrury
 
 
5 hours later…
Andre Silva
Andre Silva
1:25 PM
@gung, I was gradually doing the job of removing [best-practices] tag as per this post and recent comments beneath it, but you "rolled back" my last edit; so I wonder if I should continue carrying on this job. Tks. (/cc @Silverfish)
 
Andre Silva
Andre Silva
1:49 PM
@gung, please disregard my last comment. I missed the recent comments on that question. Tks.
 
StasK
StasK
2:05 PM
@amoeba, this is what my anti-PLS friends told me:

Maximum-likelihood PCR outperforms PLS-R (and regular PCR), particularly when measurement errors are heteroskedastic and/or intercorrelated (and when would that ever happen ;) ).

Wentzell, P.D. (2015). The Errors of My Ways: Maximum Likelihood PCA Seventeen Years after Bruce. In B.K. Lavine, S.D. Brown, & K.S. Booksh (Eds.), 40 Years of Chemometrics: From Bruce Kowalski to the Future (ACS Symposium Series, vol. 1199, pp. 31-64). American Chemical Society.
They did not have much to say whether the evils of the marketing PLS path modeling apply to chemometrics PLS regularization.
 
General Abrial
General Abrial
2:22 PM
I'm not sure why one of my recent comment flags was declined. If a mod would care to discuss, I think it would be helpful for my understanding of what to flag/not flag.
 
whuber
whuber
3:15 PM
@GeneralAbrial There does not appear to be an interface for investigating all flags, so could you indicate which post was involved and approximately when you flagged the comment?
BTW, a reason for declining a flag is usually supplied. What reason was given?
 
gung
gung
@AndreSilva, I didn't roll it back, I just added the tag. You can re-delete it, though.
I actually don't have a problem with [best-practices]. I think it's potentially meaningful (although, I can't quite come up with a definition & usage guide). Still, you can't fight the will of the people. So I guess it will go the way of the dodo.
 
amoeba
amoeba
@StasK "Maximum likelihood PCA" is another niche thing that exists only inside chemometrics, and it is even more obscure than PLS. Almost all the refs are by Wentzell. As an outsider, I am not sure I buy it.
My impression is that elastic net should outperform all of these ad hoc techniques in terms of predictive accuracy.
So I am not very interested in that fight.
 
General Abrial
General Abrial
3:36 PM
@whuber I flagged a comment in this thread. stats.stackexchange.com/questions/229165/… No decline reason was given
@whuber I flagged the post earlier today. C. 10 am
 
StasK
StasK
3:57 PM
@amoeba ha-ha @amoeba, fair enough :). I am sure each one of us have a plate full of work and fights to fight.
 
Scortchi
Scortchi
@GeneralAbrial: FYI we can't give decline reasons on comment flags (unlike answer/question flags). Don't know why - I'd like to sometimes.
 
Andre Silva
Andre Silva
4:18 PM
Yes, the backquotes was to mean something like a "roll back" (because I had only removed the tag), but not a roll back indeed as you said :); but anyway I agree to follow people's will, that is one of the reasons meta exists, I guess. On the other hand, if you would like to reactivate that meta most and discuss more about the tag, it would be ok for me, no problem (@gung).
 
General Abrial
General Abrial
@Scortchi Oh, I didn't realize that. Thanks for letting me know. It seemed like a reasonable thing to flag to me, so I'd just like to know what kinds of things I should/should not flag for moderation.
 
Scortchi
Scortchi
Well, certainly it's a reasonable thing to flag.
Thinking about it I'd probably have declined it too - perhaps a little over-heated, but there's a point to it & "trigger-happy" isn't rude.
Plus moderators sign up to take a bit of flak at least, & the user'd probably see deleting it as yet more aggravation.
 
whuber
whuber
4:34 PM
@GeneralAbrial I agree with @Scortchi. I am glad it was flagged. The only choices offered in response to a flagged comment are "delete" (which makes it disappear) and "dismiss" (which I suppose comes back to you as being declined). Since this post and the topic of the flag are actively under discussion in meta, I thought it better to keep the comment up at least for a while.
I try to be conservative about deleting anything unilaterally, so if I'm in doubt about the status of a comment I will tend to let it stand.
 
General Abrial
General Abrial
5:21 PM
@whuber @Scortchi ok, these are consistent with my surmise. when it was declined, I thought that these might be the reasons but I wanted the confirmaton. thanks!
 
 
4 hours later…
usεr11852
usεr11852
9:20 PM
I recently answered some question relating to number of goals scored in football and I noticed that we do not have any tag about sports/sports statistics in general. The terms 'football', 'soccer', 'basketball`return approximately ~350 all-together, do you think it makes sense to have a [sports statistics] tag?
BTW, apparently we have a [transportation] tag (with 5 member only)... Are we sure it is not overly specific?
 
 
1 hour later…
Kodiologist
Kodiologist
10:33 PM
We do have [games].
[sports] would probably be fine; [sports statistics] is a bit redundant.
 
amoeba
amoeba
@usεr11852 [Transportation] appears to be a pretty useless tag. I'd say feel free to remove it.
And I agree with @Kodiologist that [sports] would be fine.
 
Glen_b
Glen_b
11:29 PM
@gung I migrated this as requested but it looks close enough to stuff they often migrate to us (in spite of the fact that it would close here) that it may come back.
 
gung
gung
11:50 PM
@Glen_b, hmmm thanks. I guess we can hope for the best. Maybe it's really about understanding the underlying statistical issues. I don't know that material well enough, so someone more expert would have to make the case for reopening. I do worry that we've gotten a little too strict, but it seems people disagreed with me.
@usεr11852, what do you think the scope of the tag would be? Colloquially, a lot of people use 'statistics' in sports to just mean numbers (or more charitably, metrics of some latent performance variable). I'm not sure what the (legitimate) statistical issues would be in most of the things I've heard sports fans say in regards to 'statistics'. (Note that I'm not a sports fan, so I could easily be missing something.)
 

« first day (2185 days earlier)      last day (2823 days later) »