nLab Gaussian probability distribution

Redirected from "standard deviations".
Contents

Contents

Definition

A probability distribution on a Cartesian space ℝ n\mathbb{R}^n is called Gaussian or a normal distribution if it is of the form

μ A:x→↦detA(2π) n/2exp(−12⟨x→,Ax→⟩)dx 1⋯dx n. \mu_A \;\colon\; \vec x \mapsto \frac{\sqrt{det A}}{(2\pi)^{n/2}} \exp\left(-\tfrac{1}{2} \left\langle \vec x , A \vec x\right\rangle\right) \,d x^1 \cdots d x^n \,.

where AA is some n×nn \times n matrix such that ⟨−,A−⟩\langle -, A-\rangle is a positive definite bilinear form. Here detA\det A denotes the determinant and ⟨−,−⟩\langle -,-\rangle is the canonical bilinear form on ℝ n\mathbb{R}^n.

Since detA\sqrt{\det A} is the coordinate of the volume element vol Avol_A associated with AA, we may equivalently write this as

μ A:x→↦1(2π) n/2exp(−12⟨x→,Ax→⟩)vol A. \mu_A \;\colon\; \vec x \mapsto \frac{1}{(2\pi)^{n/2}} \exp\left(-\tfrac{1}{2} \left\langle \vec x , A \vec x\right\rangle\right) \,vol_A \,.

The mean of this distribution is 0→\vec{0}; for a distribution with mean c→\vec{c}, replace ⟨x→,Ax→⟩\langle{\vec{x},A \vec{x}}\rangle with ⟨x→−c→,Ax→−Ac→⟩\langle{\vec{x} - \vec{c},A \vec{x} - A \vec{c}}\rangle.

The matrix AA is the inverse of the covariance matrix. In particular, for n=1n = 1, we may write x 2/σ 2x^2/\sigma^2 (or (x−c) 2/σ 2(x-c)^2/\sigma^2 for mean cc) in place of ⟨x→,Ax→⟩\langle{\vec{x},A \vec{x}}\rangle, where σ\sigma is the standard deviation; similarly, detA\sqrt{\det A} becomes 1/σ1/\sigma. This gives the form

μ σ:x↦1σ2πexp(−(x−c) 22σ 2)dx\mu_\sigma \;\colon\; x \mapsto \frac{1}{\sigma \sqrt{2\pi}} \exp\left(-\frac{(x-c)^2}{2\sigma^2}\right) \, dx

In the case that n=1n = 1 one denotes this distribution by 𝒩(c,σ 2)\mathcal{N}(c,\sigma^2).

Square as a random variable

The square Y=X 2Y = X^2 of a random number with probability distribution 𝒩(0,σ 2)\mathcal{N}(0,\sigma^2) follows a scaled Chi-squared distribution.

This has positive expected value 0<𝔼[X 2]0 \lt \mathbb{E}[X^2] and bounded variance Var(X 2)<∞\text{Var}(X^2) \lt \infty.

The probability density function of Y=X 2Y = X^2 is 

f Y(y)=12πσ 2yexp(−y2σ 2),y∈(0,∞) f_Y(y) = \frac{1}{\sqrt{2\pi \sigma^2 y}} \exp\left(-\frac{y}{2\sigma^2}\right), y \in (0,\infty)

with f Y(0)=0f_Y(0) = 0. For X∼𝒩(0,σ 2)X \sim \mathcal{N}(0, \sigma^2), the general moment formulas are:

𝔼[X n]={(n−1)!!σ n=n!2 n/2(n/2)!σ n if n is even 0 if n is odd \mathbb{E}[X^n] = \begin{cases} (n-1)!! \, \sigma^n = \frac{n!}{2^{n/2}(n/2)!} \, \sigma^n & \text{if }\, n\, \text{ is even} \\ 0 & \text{if }\, n\, \text{ is odd} \end{cases}

and

Var(X n)=𝔼[X 2n]−(𝔼[X n]) 2=σ 2nVar(Z n) \text{Var}(X^n) = \mathbb{E}[X^{2n}] - (\mathbb{E}[X^n])^2 = \sigma^{2n} \text{Var}(Z^n)

where Z∼𝒩(0,1)Z \sim \mathcal{N}(0, 1) is the standard normal variable.

Specifically, for n=2n = 2, one has 𝔼[X 2]=Var(X)=σ 2\mathbb{E}[X^2] = \text{Var}(X) = \sigma^2 and Var(X 2)=𝔼[X 4]−(𝔼[X 2]) 2=3σ 4−σ 4=2σ 4\text{Var}(X^2) = \mathbb{E}[X^4] - (\mathbb{E}[X^2])^2 = 3\sigma^4 - \sigma^4 = 2\sigma^4.

If we define the standardized centered variable W=X 2−𝔼[X 2]Var(X)=X 2−σ 2σ 2W = \frac{X^2 - \mathbb{E}[X^2]}{\text{Var}(X)} = \frac{X^2 - \sigma^2}{\sigma^2}, then by linearity of expectation:

𝔼[W]=𝔼[X 2]−σ 2σ 2=σ 2−σ 2σ 2=0 \mathbb{E}[W] = \frac{\mathbb{E}[X^2] - \sigma^2}{\sigma^2} = \frac{\sigma^2 - \sigma^2}{\sigma^2} = 0

By scaling properties of variance, dividing by the constant σ 2\sigma^2 rescales the variance by 1σ 4\frac{1}{\sigma^4}:

Var(W)=Var(X 2−σ 2σ 2)=1σ 4Var(X 2)=2σ 4σ 4=2 \text{Var}(W) = \text{Var}\left(\frac{X^2 - \sigma^2}{\sigma^2}\right) = \frac{1}{\sigma^4}\text{Var}(X^2) = \frac{2\sigma^4}{\sigma^4} = 2

Substituting X=σZX = \sigma Z directly into WW confirms this independence from σ\sigma:

W=(σZ) 2−σ 2σ 2=σ 2(Z 2−1)σ 2=Z 2−1 W = \frac{(\sigma Z)^2 - \sigma^2}{\sigma^2} = \frac{\sigma^2(Z^2 - 1)}{\sigma^2} = Z^2 - 1

Because W=Z 2−1W = Z^2 - 1 identically for all 0<σ0 \lt \sigma , the standardized deviation WW is strictly scale-invariant with zero mean and variance Var(W)=2\text{Var}(W) = 2.

References

See also

Last revised on November 5, 2026 at 04:38:12. See the history of this page for a list of all contributions to it.