跳转至
\[ \newcommand{\bs}{\boldsymbol} \newcommand{\bsX}{\boldsymbol{X}} \newcommand{\bf}{\mathbf} \newcommand{\msc}{\mathscr} \newcommand{\mca}{\mathcal} \newcommand{\T}{\text{T}} \newcommand{\rme}{\mathrm{e}} \newcommand{\rmi}{\mathrm{i}} \newcommand{\rmj}{\mathrm{j}} \newcommand{\rmd}{\mathrm{d}} \newcommand{\rmm}{\mathrm{m}} \newcommand{\rmb}{\mathrm{b}} \newcommand{\and}{\land} \newcommand{\or}{\lor} \newcommand{\exist}{\exists} \newcommand{\sube}{\subseteq} \newcommand{\lr}[3]{\left#1 #2 \right#3} \newcommand{\intfy}{\int_{-\infty}^{+\infty}} \newcommand{\sumfy}[1]{\sum_{#1=-\infty}^{+\infty}} \newcommand{\vt}{\vartheta} \newcommand{\ve}{\varepsilon} \newcommand{\vp}{\varphi} \newcommand{\Var}{\text{Var}} \newcommand{\Cov}{\text{Cov}} \newcommand{\edef}{\xlongequal{def}} \newcommand{\prob}{\text{P}} \newcommand{\Exp}{\text{E}} \newcommand{\t}[1]{\text#1} \newcommand{\N}{\mathbb{N}} \newcommand{\Z}{\mathbb{Z}} \newcommand{\Q}{\mathbb{Q}} \newcommand{\R}{\mathbb{R}} \newcommand{\C}{\mathbb{C}} \newcommand{\versionofnewcommand}{\text{260125}} \]

\(\mathrm{II.\ 2.}\) Basic Probabiltiy Theory

14. Probability Triple

Definition:

By probability triple, we mean a measure space \((\Omega,\msc{F},\prob)\) of total mass \(\prob(\Omega)=1\). An element \(E\) of \(\msc{F}\) will be called an event, and \(\prob(E)\) will be called the probability of the event \(E\). A statement \(S\) about outcomes \(\omega\) in \(\Omega\) is said to be true almost surely (a.s.) if

\[ F = \{ \omega \ |\ S(\omega)\text{ is true} \} \in \msc{F}\quad \text{and}\quad \prob(F)=1 \]

15. First Borel-Cantelli Lemma

Definition:

Suppose that \(\{E_n\}\) is a sequence of events. We define

\[ \begin{aligned} \lim\sup E_n =\ & \bigcap_{m} \bigcup_{n \geq m} E_n\\ = &\ \{ \omega \ |\ \text{for every } m,\ \exist \ n(\omega)\geq m \text{ such that }\omega \in E_{n(\omega)} \}\\ = &\ \{ \omega \ |\ \omega\in E_n \text{ for infinitely many n} \} \end{aligned} \]

Lemma (Reverse Fatou Lemma for Sets):

\[ \prob(\lim\sup E_n) \geq \lim\sup \prob(E_n) \]

Lemma (First Borel-Cantelli Lemma):

Let \(\{E_n\}\) be a sequence of events such that \(\sum_n \prob(E_n)<\infty\). Then,

\[ \prob(\lim\sup E_n) = \prob(E_n, \text{i.o.}) = 0 \]

16. Random Variable; Distribution Function

Definition:

Let \((E,\msc{E})\) be a measurable space. By an \((E,\msc{E})\)-valued random variable \(X\) carried by our probability triple \((\Omega,\msc{F},\prob)\), we mean an \((\msc{F/E})\)-measurable map \(X\) from \(\Omega\) to \(E\), so that \(X^{-1}:\msc{E} \to \msc{F}\). By the law \(\Lambda_X\) of \(X\), we mean the probability measure \(\Lambda_X = P \circ X^{-1}\) on \((E,\msc{E})\) so that

\[ \Lambda_X(A) = \prob(X \in A) = \prob(\omega \ |\ X(\omega)\in A)\quad(A\in\msc{E}) \]

Suppose that for \(i=1,2\), \((E_i,\msc{E}_i)\) is a measurable space and that \(X_i\) is an \((E_1,\msc{E}_i)\)-valued random variable. Let

\[ E = E_1 \times E_2,\quad \msc{E}=\msc{E}_1 \times \msc{E}_2 \]

and

\[ X(\omega) = (X_1(\omega),X_2(\omega)) \in E \]

We say that the joint law \(\Lambda_{X_1,X_2}\) of \(X_1\) and \(X_2\) is then defined to be the law of \(X\).

If the random variable \(X\) is \(\R\)-valued, then it follows from the Uniqueness Lemma and the fact that \(\msc{B}=\sigma(\pi(\R))\), where

\[ \pi(\R)\edef \{(-\infty,x]:x\in\R\}, \]

that the law of \(X\) is determined by the distribution function \(F_X\) of \(X\):

\[ F_X(x) = \prob(X \in (-\infty,x])=\prob(X \leq x),\quad x\in\R \]

Inversely we can say that the distribution function is a special law of random variables.

17. Expectation