2
$\begingroup$

Let's $x$ be a random variable with a binomial distribution ($x \sim B(n,p)$). I know that the expected value of a binomial is $E(x) = n \cdot p $ but the inverse of a binomial?

  • $E\big(\frac{1}{x}\big)$ = ???

EDIT

$Y \sim unknown$ and $X \sim B(n,p)$

$E\big(\frac{Y}{X}\big) = E(Y) \cdot E\big(\frac{1}{X}\big)$

$\endgroup$
6
  • 1
    $\begingroup$ I wonder whether such general result exists! $\endgroup$
    – hola
    Jun 27, 2014 at 14:44
  • 3
    $\begingroup$ What is $1/x$ when $x=0$? $\endgroup$ Jun 27, 2014 at 14:50
  • $\begingroup$ @pushpen.paul I don't know if such a general result exist but since i need to compute the expected value of a ratio between two random variables I found this problem. $\endgroup$
    – gmeroni
    Jun 27, 2014 at 15:17
  • 1
    $\begingroup$ In that case, the expectation is obviously infinite, since there is positive probability that $x=0$. $\endgroup$ Jun 27, 2014 at 15:36
  • 4
    $\begingroup$ It takes only a single infinite case to make the expectation infinite! $\endgroup$
    – whuber
    Jun 27, 2014 at 15:36

1 Answer 1

2
$\begingroup$

Since $\Pr[X = 0] = p^n > 0$, what is $${\rm E}\left[\frac{1}{X}\right] = \sum_{i=0}^n \frac{1}{i} \Pr\left[X = i\right] = ?$$

$\endgroup$
1
  • 2
    $\begingroup$ Check your expression carefully. Should $i$ start from $0$? Doesn't it make the whole expression meaningless? $\endgroup$
    – hola
    Jun 29, 2014 at 11:00

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.