Talk:Scott's rule
Appearance
(Redirected from Talk:Scott's Rule)
This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||||||||||
|
The intro is not very clear or accessible
[edit]The intro is not very clear or accessible, at least for someone familiar with a lot of math, but not a lot of statistics. For example, why does the dx come at the beginning of the integral defining IMSE, and are the data points drawn independently from the distribution with pdf f? Also, the intro discusses "Taylor expanding around h" in reference to an equation which doesn't contain h. It also claims that from the Taylor expansion, you can solve for the optimal h --- should this say solving for an approximately optimal h? 4Aleph4Omega4 (talk) 01:33, 5 September 2024 (UTC)