forked from emsbach/probability-and-computing-solutions
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathexercise01.20.tex
18 lines (18 loc) · 1.18 KB
/
exercise01.20.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
\paragraph{Exercise 1.20} $~$ \\[0.2cm]
\textbf{Theorem 1.10}: If $E_1, E_2, ..., E_n$ are mutually independent, then so
are $\bar{E_1}, \bar{E_2}, ..., \bar{E_n}$. \\[0.2cm]
\textit{Proof}. Let $E_1, E_2, ..., E_n$ be mutually independent. Then, for any
subset $I \subseteq [1, n]$,
\begin{align*}
\pr\left(\bigcap_{i \in I} \bar{E_i}\right)
&= \pr\left(\bar{\bigcup_{i \in I} E_i}\right) \\
&= 1 - \pr\left(\bigcup_{i \in I} E_i\right) \\
&= 1 - \left( \sum_{i \in I}\pr(E_i) - \sum_{i,j \in I, i < j} \pr(E_i \cap E_j) + ... + (-1)^{|I| + 1}\pr\left(\bigcap_{i\in I}E_i\right) \right) \\
&= 1 - \left( \sum_{i \in I} \pr(E_i) - \sum_{i,j \in I, i < j} \pr(E_i) \cdot \pr(E_j) + ... + (-1)^{|I| + 1}\prod_{i\in I} \pr(E_i) \right) \\
&= 1 - \sum_{i \in I}\pr(E_i) + \sum_{i,j \in I, i < j} \pr(E_i) \cdot \pr(E_j) + ... + (-1)^{|I|}\prod_{i\in I} \pr(E_i) \\
&= \prod_{i \in I}\left(1 - \pr(E_i)\right) \\
&= \prod_{i \in I}\pr(\bar{E_i}).
\end{align*}
We have used De Morgan's law in the first line, the inclusion-exclusion principle
in the third line and the assumption that $E_1, E_2, ..., E_n$ are mutually
independent combined with definition 1.3 in the fourth line.