Timeline for Why does component-wise median not make sense in higher dimensions?
Current License: CC BY-SA 4.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 13, 2020 at 10:42 | history | edited | Alex | CC BY-SA 4.0 |
added 2 characters in body
|
| Apr 13, 2020 at 4:50 | comment | added | Sebastian | You could Look at the concept of a half-mass space | |
| Apr 10, 2020 at 20:43 | vote | accept | Alex | ||
| Apr 9, 2020 at 0:25 | history | became hot network question | |||
| Apr 8, 2020 at 21:12 | answer | added | whuber♦ | timeline score: 25 | |
| Apr 8, 2020 at 21:00 | history | tweeted | twitter.com/StackStats/status/1247992710457417728 | ||
| Apr 8, 2020 at 18:40 | comment | added | Alex | @user20160 thanks a lot, you should post that as an answer! | |
| Apr 8, 2020 at 17:21 | comment | added | user20160 | What you're describing (the vector of component-wise medians) is called the marginal median. There are various ways to generalize the median to multivariate settings, including the marginal median, the L1 median, and the center point. Each of these reduce to the standard median in 1d, but have different properties in higher dimensions. I wouldn't say that the marginal median doesn't make sense. In fact, people have used it for various applications. Rather, it depends on what properties matter in your particular problem. | |
| Apr 8, 2020 at 16:42 | history | edited | Alex | CC BY-SA 4.0 |
added 202 characters in body
|
| Apr 8, 2020 at 16:25 | review | First posts | |||
| Apr 8, 2020 at 18:07 | |||||
| Apr 8, 2020 at 16:21 | history | asked | Alex | CC BY-SA 4.0 |