Re: multivariate Gaussian
- From: Jack Tomsky <jtomsky@xxxxxxxxxxxxx>
- Date: Wed, 26 Sep 2007 11:44:28 EDT
hello all, suppose X is a n dimensional 0 mean normal
with variance in
each direction sigma^2. now assume a n dimensional
ball of radius R
centered at 0. is it possible to get an analytical
expression for R
such that the probability mass inside the ball=0.5?
thanks
R(0.5) is sigma times the square-root of the median of a chi-square with n degrees of freedom. For n = 2, it is exactly sigma*sqrt(2*ln(2)). For n > 2, you might use analytic approximations for the median of a chi-square.
Jack
.
- References:
- multivariate Gaussian
- From: Kaushik
- multivariate Gaussian
- Prev by Date: Re: How do I express the regression function in the original units?
- Next by Date: Chisquare distribution
- Previous by thread: Re: multivariate Gaussian
- Next by thread: Is having an equal number of replications an advantage? (Question on Lack of Fit test in Simple Linear Regression)
- Index(es):
Relevant Pages
|
|
