E[X^4] = 3σ^4

Tags:

2024.05.08


(dWt)2=dt(dW_t)^2=dt 인 이유에 대해 보던 중, 확률변수 X∼N(0,σ2)X \sim \mathcal{N}(0,\sigma^2)에 대해 E[X4]=3σ4\mathbb{E}[X^4]=3\sigma^4 인 이유를 stackexchange까지 찾아봐도 상세한 계산까지 설명된 건 없어서 직접 계산을 해봐야 했다.

E[X4]:=∫−∞∞x4φ(x)dx=∫−∞∞x412πσe−x22σ2dx(pdf of normal distribution)=∫−∞∞x3(σ⋅x2πσ2e−x22σ2)dx=[x3⋅(−σ2πσe−x22σ2)]−∞∞−∫−∞∞3x2(−σ2πe−x22σ2)dx(integration by parts)=0−∫−∞∞3x2(−σ22πσe−x22σ2)dx=3σ2∫−∞∞x2(12πσe−x22σ2)dx=3σ2⋅σ2(definition of variance)=3σ4\begin{aligned} \mathbb{E}[X^4] & := \int_{-\infty}^{\infty}{x^4\varphi(x)}dx \\ & = \int_{-\infty}^{\infty}{x^4 \textstyle \frac{1}{\sqrt{2\pi}\sigma}e^{-\frac{x^2}{2\sigma^2}}}dx && \text{(pdf of normal distribution)}\\ & = \int_{-\infty}^{\infty}{x^3 (\textstyle \frac{\sigma \cdot x}{\sqrt{2\pi}\sigma^2}e^{-\frac{x^2}{2\sigma^2}})}dx \\ & = \left[ x^3 \cdot (\textstyle -\frac{\sigma}{\sqrt{2\pi}\sigma}e^{-\frac{x^2}{2\sigma^2}}) \right] _{-\infty} ^{\infty} - \int_{-\infty}^{\infty}{3x^2 (\textstyle -\frac{\sigma}{\sqrt{2\pi}}e^{-\frac{x^2}{2\sigma^2}})}dx && \text{(integration by parts)}\\ & = 0 - \int_{-\infty}^{\infty}{3x^2 (\textstyle -\frac{\sigma^2}{\sqrt{2\pi}\sigma}e^{-\frac{x^2}{2\sigma^2}})}dx \\ & = 3\sigma^2 \int_{-\infty}^{\infty}{x^2 (\textstyle \frac{1}{\sqrt{2\pi}\sigma}e^{-\frac{x^2}{2\sigma^2}})}dx \\ & = 3\sigma^2 \cdot \sigma^2 && \text{(definition of variance)} \\ & = 3\sigma^4 \end{aligned}

나아가서 E[X2n]\mathbb{E}[X^{2n}] 에 대한 일반화도 가능하다.

E[X2n]:=∫−∞∞x2nφ(x)dx=∫−∞∞x2n−1(σ⋅x2πσ2e−x22σ2)dx=[x2n−1⋅(−σ2πσe−x22σ2)]−∞∞−∫−∞∞(2n−1)x2n−2(−σ2πe−x22σ2)dx=0−∫−∞∞(2n−1)x2n−2(−σ22πσe−x22σ2)dx=(2n−1)σ2∫−∞∞x2n−2(12πσe−x22σ2)dx=(2n−1)σ2 E[X2n−2]=(2n−1)σ2⋅(2n−3)σ2 E[X2n−4]⋮=(2n−1)!!σ2n\begin{aligned} \mathbb{E}[X^{2n}] & := \int_{-\infty}^{\infty}{x^{2n}\varphi(x)}dx \\ & = \int_{-\infty}^{\infty}{x^{2n-1} (\textstyle \frac{\sigma \cdot x}{\sqrt{2\pi}\sigma^2}e^{-\frac{x^2}{2\sigma^2}})}dx \\ & = \left[ x^{2n-1} \cdot (\textstyle -\frac{\sigma}{\sqrt{2\pi}\sigma}e^{-\frac{x^2}{2\sigma^2}}) \right] _{-\infty} ^{\infty} - \int_{-\infty}^{\infty}{(2n-1)x^{2n-2} (\textstyle -\frac{\sigma}{\sqrt{2\pi}}e^{-\frac{x^2}{2\sigma^2}})}dx \\ & = 0 - \int_{-\infty}^{\infty}{(2n-1)x^{2n-2} (\textstyle -\frac{\sigma^2}{\sqrt{2\pi}\sigma}e^{-\frac{x^2}{2\sigma^2}})}dx \\ & = (2n-1)\sigma^2 \int_{-\infty}^{\infty}{x^{2n-2} (\textstyle \frac{1}{\sqrt{2\pi}\sigma}e^{-\frac{x^2}{2\sigma^2}})}dx \\ & = (2n-1)\sigma^2 \ \mathbb{E}[X^{2n-2}] \\ & = (2n-1)\sigma^2 \cdot (2n-3)\sigma^2 \ \mathbb{E}[X^{2n-4}] \\ & \qquad \vdots \\ & = (2n-1)!!\sigma^{2n} \end{aligned}

+ 이거 쓰다가 알았는데 일반적인 Factorial(!) 말고 Double Factorial(!!)도 존재한다.
e.g.7!!=7⋅5⋅3⋅1\quad e.g. 7!! = 7 \cdot 5 \cdot 3 \cdot 1

 

2024.05.08
2024.05.10