Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. 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

$$\sum_{n=0}^\infty \binom{n}{k}p^k(1-p)^{n-k}\frac{\lambda^ne^{-\lambda}}{n!} = \frac{(\lambda p)^ke^{-\lambda p}}{k!}$$

That is, given a random variable $X$ with Poisson distribution $X \sim \operatorname{Poisson}(\lambda)$, and, given $X = n$, random variable $Y$ is distributed by a binomial such that $Y \sim B(n,p): p \in [0,1]$. Given $n$, $P(Y = k|X = n) = \binom{n}{k}p^k(1-p)^{n-k}$. Given $\lambda$, $P(X=n) = \dfrac{\lambda^ne^{-\lambda}}{n!}$. Therefore, $P(Y = k|\lambda) = \sum_{n=0}^\infty P(Y = k | X = n)P(X = n | \lambda) = \sum_{n=0}^\infty \binom{n}{k}p^k(1-p)^{n-k}\dfrac{\lambda^ne^{-\lambda}}{n!}$. What properties and identities can be used to demonstrate the equality above?

share|cite|improve this question
up vote 2 down vote accepted

A start: We want $\Pr(Y=k)$, so we should end up with an expression that involves $k$. Let $m=n-k$. Effectively we are summing from $m=0$ to $\infty$.

Take the terms that involve only $k$, $\lambda$, and $p$ to the outside, not forgetting that $\lambda^n=\lambda^{m+k}=\lambda^k\lambda^m$ and $\binom{m+k}{k}\cdot \frac{1}{(m+k)!}=\frac{1}{k!}\cdot\frac{1}{m!}$. The stuff that remains on the inside looks like $$\sum_{m=0}^\infty \frac{(\lambda(1-p))^m}{m!}.$$ From the familiar series expansion of $e^x$, we see that this sum is $e^{\lambda(1-p)}$.

Remark: The above approach is crudely computational. A much better conceptual justification has been given in the answer by Robert Israel.

share|cite|improve this answer
Perfect. This is exactly what I was looking for. – brian Jul 30 '12 at 19:31
...on the other hand, one should know how to do "crude" computations. Sometimes that's all you've got. – Michael Hardy Jul 30 '12 at 23:09

Consider a "compound Poisson process": you have a Poisson process with rate $\lambda$, but each event, independently, is type A with probability $p$. Let $Y$ be the number of events of type A in the time interval $[0,1]$. Then $Y$ has a Poisson distribution with parameter $\lambda p$. This can be seen because the type A events form a Poisson process with rate $\lambda p$.

share|cite|improve this answer

Your Answer


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.