Timeline for What is the problem with $p > n$?
Current License: CC BY-SA 4.0
13 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 6, 2019 at 0:00 | history | tweeted | twitter.com/StackStats/status/1081702005964460034 | ||
| Jan 5, 2019 at 13:46 | vote | accept | EconBoy | ||
| Jan 5, 2019 at 13:30 | answer | added | boomkin | timeline score: 3 | |
| Jan 5, 2019 at 12:47 | history | edited | Frank Harrell |
edited tags
|
|
| Jan 5, 2019 at 12:46 | answer | added | Frank Harrell | timeline score: 12 | |
| Jan 5, 2019 at 11:43 | comment | added | nope | Sometimes it just is not feasible to have enough observations to counter all the variables that need to be accounted for. Classic examples are medical studies, where the number of people you can treat is very limited. | |
| Jan 5, 2019 at 10:15 | history | edited | kjetil b halvorsen♦ | CC BY-SA 4.0 |
added 2 characters in body; edited title
|
| Jan 5, 2019 at 10:14 | comment | added | kjetil b halvorsen♦ | There might be cases where, once you got a subject in your sample, it is cheap to measure (tens of) thousands of variables---at least not more expensive than 10 or 100. Bioinformatics? automated sensors? | |
| Jan 5, 2019 at 5:03 | history | edited | EconBoy | CC BY-SA 4.0 |
[Edit removed during grace period]
|
| Jan 5, 2019 at 5:00 | comment | added | EconBoy | Yes, no doubt I don't. I get the concept part but partly about the combination of the model in math. why would we suffer from p > n? | |
| Jan 5, 2019 at 4:53 | comment | added | Alexis | I suspect you have not yet been enlightened about step-wise model selection, but I like this question. | |
| Jan 5, 2019 at 4:49 | history | edited | Alexis | CC BY-SA 4.0 |
deleted 13 characters in body; edited title
|
| Jan 5, 2019 at 4:44 | history | asked | EconBoy | CC BY-SA 4.0 |