高斯分布概率密度函数积分推导
高斯分布:
$f(x) = \frac{1}{\sqrt{2\pi }\sigma }exp(-\frac{(x-\mu)^{2}}{2\sigma^{2}})$
标准高斯分布:
$f(x) = \frac{1}{\sqrt{2\pi } }exp(-\frac{x^{2}}{2})$
一个高斯分布只需线性变换即可化为标准高斯分布,所以只需推导标准高斯分布概率密度的积分。由:
$\int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi } }exp(-\frac{x^{2}}{2})dx = \int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi } }exp(-\frac{y^{2}}{2})dy$
得:
$\int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi } }exp(-\frac{x^{2}}{2})dx\int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi } }exp(-\frac{y^{2}}{2})dy = \frac{1}{2\pi}\int_{-\infty }^{+\infty }\int_{-\infty }^{+\infty }exp(-\frac{(x^{2}+y^{2})}{2})dxdy$
变换为极坐标得:
$x = rcos\theta , y = rsin\theta$
$\frac{1}{2\pi}\int_{-\infty }^{+\infty }\int_{-\infty }^{+\infty }exp(-\frac{(x^{2}+y^{2})}{2})dxdy = \frac{1}{2\pi}\int_{0}^{2\pi }\int_{0}^{+\infty }exp(-\frac{r^{2}}{2})rdrd\theta$
$\frac{1}{2\pi}\int_{-\infty }^{+\infty }\int_{-\infty }^{+\infty }exp(-\frac{(x^{2}+y^{2})}{2})dxdy = \int_{0}^{+\infty }exp(-\frac{r^{2}}{2})rdr$
令$t = \frac{r^{2}}{2} $,得:
$\int_{0}^{+\infty }exp(-\frac{r^{2}}{2})rdr = \int_{0}^{+\infty }exp(-t)dt = 1$
终得:
$\int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi } }exp(-\frac{x^{2}}{2})dx = \int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi } }exp(-\frac{y^{2}}{2})dy = 1$
由于标准高斯分布概率密度函数$f(x)$满足:
$\int_{-\infty }^{+\infty }f(x)dx = \int_{-\infty }^{+\infty }f(x-c)dx$
$\int_{-\infty }^{+\infty }f(x)dx = \frac{1}{a}\int_{-\infty }^{+\infty }f(\frac{x}{a})dx$
所以任意高斯分布概率密度函数积分为1。
高斯分布的均值:
$ E(x) = \int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi }\sigma }exp(-\frac{(x-\mu)^{2}}{2\sigma^{2}})xdx $
$= \int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi }\sigma }exp(-\frac{(x-\mu)^{2}}{2\sigma^{2}})(x-\mu)dx + \mu\int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi }\sigma }exp(-\frac{(x-\mu)^{2}}{2\sigma^{2}})dx $
由奇偶性得:
$ E(x) = \mu\int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi }\sigma }exp(-\frac{(x-\mu)^{2}}{2\sigma^{2}})dx = \mu$
高斯分布的方差:
$Var(x) = \int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi }\sigma }exp(-\frac{(x-\mu)^{2}}{2\sigma^{2}})(x-\mu)^{2}dx$
$\int_{-\infty }^{+\infty }\frac{1}{\sqrt{2\pi }\sigma }exp(-\frac{y^{2}}{2\sigma^{2}})y^{2}dy = \sigma^{2} $