Talk:Sub-Gaussian distribution
Add topicLatest comment: 6 months ago by ~2026-65071-7 in topic Variance proxy bounds
 | This article is rated C-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | ||||||||||
| |||||||||||
Variance proxy bounds
[edit]There is an issue with the variance proxy bounds. It says that the bernoulli distribution with 1/2 weights on (-1/2) and (1/2) has variance proxy of 1, but the bound on general distributions on the interval [a,b] are tighter. In this case this (b-a)= ((1/2 - (-1/2)) = 1 and consequently the vp norm would be 1/4. This conflicts with the claim that the vp is 1. I don't know which is right — Preceding unsigned comment added by ~2026-65071-7 (talk) 07:38, 30 January 2026 (UTC)
[Untitled]
[edit]it is a boring concept.— Preceding unsigned comment added by 154.5.98.34 (talk • contribs) 17:05, 30 November 2019 (UTC)
- So is the successor function in the theory of ordinals. Dull as dishwater. 2A02:1210:2642:4A00:BC9C:833A:6804:E156 (talk) 13:04, 29 July 2023 (UTC)
Proof of (3) => (2) - use of monotone convergence theorem
[edit]Where is the monotone convergence theorem used here? B52481 (talk) 23:59, 13 June 2023 (UTC)
- It looked like series only with no special convergence issues. I removed the invocation. 2A02:1210:2642:4A00:BC9C:833A:6804:E156 (talk) 13:04, 29 July 2023 (UTC)