Re: Chisquare distribution
- From: Jack Tomsky <jtomsky@xxxxxxxxxxxxx>
- Date: Wed, 26 Sep 2007 16:19:03 EDT
hello all,
for a chi-square distribution of n degrees of freedom
the mean value
is n. it is clear that it can not be said that if X
is distributed as
chi-square with degrees of freedom n then
approximately half of the
probability mass lies in the region {x: 0<=x<=n}.
however, is it possible to get any close estimate a
as a function of n
i.e. a(n) such that for the region {x:0<=x<=a(n)} the
probability mass
is atmost half?
thanks
An approximation for the median of a chi-square when the degrees of freedom n is large, say n > 30, is
n*[1-2/(9n)]^3.
Jack
.
- References:
- Chisquare distribution
- From: Kaushik
- Chisquare distribution
- Prev by Date: Solutions Manual
- Next by Date: problem with logistic regression
- Previous by thread: Chisquare distribution
- Next by thread: Multiplication of two Gaussians
- Index(es):
Relevant Pages
|
|
