Timeline for Expected value of subset of variables in Bayesian setting
Current License: CC BY-SA 4.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 12, 2021 at 21:04 | vote | accept | yrx1702 | ||
| Jan 12, 2021 at 21:04 | history | undeleted | yrx1702 | ||
| Jan 12, 2021 at 20:59 | history | deleted | yrx1702 | via Vote | |
| Jan 12, 2021 at 20:45 | answer | added | Xi'an | timeline score: 1 | |
| Jan 12, 2021 at 20:44 | comment | added | jld | maybe I'm not following, but it sounds like you want something of the form $\text E[(X_{(N)}, \dots, X_{(N-k)})]$ for some $k\geq 0$, where $X_{(i)}$ denotes the $i$th order statistic of the sample $(X_1,\dots,X_N)$. If they're easy to sample from why not just generate $M$ realizations of $(X_1, \dots, X_N)$ for some large $M$, sort them, take the top $k$, and average the resulting vectors? | |
| Jan 12, 2021 at 20:26 | history | asked | yrx1702 | CC BY-SA 4.0 |