I've just started working through Varadhan's Probability lecture notes, and I was wondering if there's any intuitive connection between the variance formula and Holder's inequality/ $L^p$ norms in general.

In a sense, for a probability space $(\Omega, \cal B, P)$ with $P(\Omega) = 1$ and a random variable $X\in L^2(\Omega)$, then if $X\geq 0$ we can write:

$E[X^2] - (E[X])^2 = (\int X(w)^2 dP) - (\int X(w) dP)^2 = \vert\vert X \vert\vert^2_{L^2(\Omega)} - \vert\vert X^{\frac{1}{2}} \vert\vert^4_{L^{2}(\Omega)}\geq 0 $

Is there any way to think of variance in terms of the $L^p$ norms? It's probably not a very good question, but I was hoping to make a connection between this and my Lebesgue integration/measure theory course I just finished.

  • $\begingroup$ I just wish to make a comment. Nice user name given your topic and just finished course. $\endgroup$ – Mark L. Stone May 25 '15 at 0:33

The even $p$ norms corresponds directly to the $p$th moment of the random variable. $$E[X^p] = ||X||_p^p$$ The association for odd $p$ is not as clear because the definition of the norm requires taking the absolute value of the outcomes. At best you can note: $$E[|X|^p] = ||X||_p^p$$ In this way, for a symmetric distribution, you can essentially say its characteristic function is defined by the norms of a random variable with that distribution. For a random variable that is non-negative, as you mentioned, all the norms give the characteristic function.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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