👏

Poisson distribution

2023/12/17に公開

The Poisson distribution Po(X=x|\mu) is

Po(X=x|\mu) = e^{-\mu} \frac{\mu^x}{x!}

where x \in \mathbb{N} , \mu > 0 is a constant.


The Poisson distribution can be derived from the limit of a binomial distribution.

\begin{align*} \lim_{n \rightarrow \infty, np = \mu} Bin(x=X|n, p) &= \lim_{n \rightarrow \infty, np = \mu} \binom{n}{x} p^x (1-p)^{n-x} \\ &= \lim_{n \rightarrow \infty} \frac{n(n-1)\cdots(n-x+1)}{x!} (\frac{\mu}{n})^x (1-\frac{\mu}{n})^{n-x} \\ &= \lim_{n \rightarrow \infty} \frac{n^x(1-\frac{1}{n})\cdots(1-\frac{x-1}{n})}{x!} (\frac{\mu^x}{n^x}) (1-\frac{\mu}{n})^{n-x} \\ &= \lim_{n \rightarrow \infty} \frac{(1-\frac{1}{n})\cdots(1-\frac{x-1}{n})}{x!} \mu^x (1-\frac{\mu}{n})^{n-x} \\ &= \lim_{n \rightarrow \infty} \frac{(1-\frac{1}{n})\cdots(1-\frac{x-1}{n})}{x!} \mu^x (1+\frac{1}{-\frac{n}{\mu}})^{-\frac{n}{\mu}(-\mu)} (1-\frac{\mu}{n})^{-x}\\ &= \frac{\mu^x}{x!} e^{-\mu} \\ &= Po(X=x|\mu) \end{align*}

The moment generating function of Po(X=x|\mu) is

\begin{align*} M(\theta) &= E[e^{\theta X}] \\ &= \sum_{x} e^{\theta x} e^{-\mu} \frac{\mu^x}{x!} \\ &= e^{-\mu} \sum_{x} e^{\theta x} \frac{\mu^x}{x!} \\ &= e^{-\mu} \sum_{x} \frac{(e^\theta \mu)^x}{x!} \\ &= e^{-\mu} e^{e^\theta \mu} \end{align*}
\begin{align*} M'(\theta) &= e^{-\mu} e^{e^\theta \mu} (e^\theta \mu)' \\ &= \mu e^{-\mu} e^{e^\theta \mu + \theta} \\ M''(\theta) &= \mu e^{-\mu} e^{e^\theta \mu + \theta} (e^\theta \mu + \theta)' \\ &= \mu e^{-\mu} e^{e^\theta \mu + \theta} (e^\theta \mu + 1) \end{align*}
\begin{align*} M'(0) &= \mu \\ M''(0) &= \mu e^{-\mu} e^{\mu} (\mu + 1) \\ &= \mu (\mu + 1) \end{align*}

therefore, the mean is

\begin{align*} E[X] = \mu \end{align*}

the variance is

\begin{align*} V[X] &= E[X^2] - E[X]^2 \\ &= M''(0) - M(0)^2 \\ &= \mu \end{align*}

Discussion