The distribution: $P(x)=(\lambda / \pi)^{1/2} e^{- \lambda (x-a)^2}$

Using a definition for the expectation value of something, I set up the problem like this:

$$<x>=\sum_{x=0}^\infty xP(x)$$ $$ <x>=\sum_{x=0}^\infty x(\lambda / \pi)^{1/2} e^{- \lambda (x-a)^2} $$

After simplifying the summation, I arrived at:

$$ <x>= (\lambda / \pi)^{1/2} e^{-a^2 \lambda} \sum_{n=1}^\infty e^{-(n)^2} e^{2an} $$

But thinking over this, it doesn't make intuitive sense.

  1. I solved for the constant $ (\lambda / \pi)^{1/2} $ using an integral that went from $-\infty$ to $+\infty$ (not shown, this was in an earlier part of the problem). Therefore, shouldn't my summation do the same?

  2. I don't understand what the expectation value of a gaussian distribution even is. Whatever I get at the end should tell me something - what is that something?


For the first question, yes, the expectation for a continuous distribution is indeed an integral and not a sum, i.e. in your notation it would be:

$$ \langle X \rangle = \int_{-\infty}^{+\infty} xP(x)dx $$

For the second question, the expectation of a Gaussian distribution is very intuitive since it also equal to its median and mode. In particular, if you plot the density of a given Gaussian, then its expectation is the position at which the density gets maximized! (In your case it gets maximized at $x=a$, which is the expectation).


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.