MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Given a series of samples $y_1, y_2, \ldots, y_n \sim N(0,\Sigma)$, I'm looking to find the expection, $E(y^T y)$. It's fairly easy to show that $y^T \Sigma^{-1} y \sim \chi^2$, but then I get stuck. Maybe there isn't a closed form solution for $E(y^T y)$?

Thanks

Edited: Removed expectation around product.

share|cite|improve this question
up vote 3 down vote accepted

There isn't a closed form solution for the distribution, but there is for the expectation. It's called a Quadratic Form: http://en.wikipedia.org/wiki/Quadratic_form_(statistics)

$E(y^T y) = tr(\Sigma)$

Also, I think you mean to say that $y^T \Sigma^{-1} y \sim \chi ^2$, not $E(y^T \Sigma^{-1} y) \sim \chi ^2 $

share|cite|improve this answer
    
Thanks for the pointer! Never heard of quadratic forms before - some interesting results. – James Philbin Apr 27 '11 at 17:43

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.