👌

Law of large numbers

2023/12/30に公開

We can express the law of large numbers as follows:

\begin{align*} \lim_{n \rightarrow \infty} \frac{x}{n} = p \end{align*}

where x is the number of events, n is the number of trials, and p is the probability of the event.


In the situation n \rightarrow \infty , we can consider

\begin{align*} \mathrm{Bin}(X=x | n, p) &\approx \mathcal{N}(X=x|\mu, \sigma^2) \\ \mu &\approx np \\ \sigma^2 &\approx npq \end{align*}

where q = 1 - p.


Let \bar{X} = \frac{X}{n} , then n\bar{X} = X

\begin{align*} \int f_{\bar{X}}(\bar{x}) d\bar{x} &= \int f_{X}(x) dx \\ &= \int f_{X}(n\bar{x}) n d\bar{x} \\ \end{align*}

therefore,

\begin{align*} f_{\bar{X}}(\bar{x}) &= n f_{X}(n\bar{x}) \\ &= n \mathcal{N}(X = n\bar{x}|\mu, \sigma^2) \\ &= n \frac{1}{\sqrt{2 \pi \sigma^2}} \exp \left\{ -\frac{(n \bar{x} - \mu)^2}{2 \sigma^2} \right\} \\ &= n \frac{1}{\sqrt{2 \pi npq}} \exp \left\{ -\frac{(n \bar{x} - np)^2}{2 npq} \right\} \\ &= \frac{1}{\sqrt{2 \pi} \sqrt{\frac{pq}{n}}} \exp \left\{ -\frac{(\bar{x} - p)^2}{2 \frac{pq}{n}} \right\} \\ &= \mathcal{N}(\bar{X}=\bar{x}|\mu_{\bar{X}} = p, \sigma_{\bar{X}}^2 = \frac{pq}{n}) \end{align*}

when n \rightarrow \infty, \sigma_{\bar{X}} \rightarrow 0 and \bar{x} = \frac{x}{n} \rightarrow p.

Discussion