11
$\begingroup$

One of the consequences of the axioms of probability ($\sigma$ field and probability axiom) is the "infinite subset" and "infinite union" property, I can't figure out how it follows from them.

if $A_1 \subseteq A_2 \subseteq \cdots$ then

$$P\left(\bigcup_{i=1}^{\infty}A_i \right) = \lim_{k \to \infty} P(A_k) $$

If $A_1 \supseteq A_2 \supseteq \cdots$ then

$$P\left(\bigcap_{i=1}^{\infty}A_i \right) = \lim_{k \to \infty} P(A_k) $$

$\endgroup$
3
  • $\begingroup$ And... what is it exactly that escapes you? $\endgroup$
    – Did
    Commented Aug 18, 2014 at 20:21
  • $\begingroup$ I think you need to use the fact that probability is an additive function on the sets defined by the sample space. $\endgroup$
    – John
    Commented Aug 18, 2014 at 20:32
  • $\begingroup$ see a detailed proof here: randomservices.org/random/prob/Measure.html $\endgroup$
    – Yibo Yang
    Commented Jul 26, 2017 at 4:13

1 Answer 1

15
$\begingroup$

The properties that you mention in your question are traditionally referred to as "continuity from above" and "continuity from below" respectively (see the math.stackexchange question Continuity from below and above for a justification of this terminology).

As for understanding how these properties follow from $\sigma$-field axioms such as

Countable additivity: if $\{E_n:n\in\mathbb N\}$ are disjoint events, then $$\Pr\left[\bigcup_nE_n\right]=\sum_n\Pr[E_n]=\lim_{n\to\infty}\sum_{i=1}^n\Pr[E_i],$$

one of the best ways to get an idea of how to proceed is to look at drawings of sets.

Let's look at the example where $A_1\subset A_2\subset A_3\subset A_4\subset\cdots$ (continuity from below). One can represent these sets graphically as (only $A_1,A_2,A_3$ are drawn): Drawing1

Suppose that we define sets $B_1,B_2,B_3,\ldots$ as $$B_1=A_1\\ B_2=A_2\setminus A_1\\ B_3=A_3\setminus(A_1\cup A_2)\\ \ldots\\ B_n=A_n\setminus(A_1\cup\cdots\cup A_{n-1})\\ \ldots,$$ and so on. Graphically, this would look like: enter image description here

Then, it seems intuitive enough that the union of every $A_i$ up to $n$ and the union of every $B_i$ up to $n$ are the same for every $n$, and that the $B_i$ are disjoint (though one still needs to demonstrate this rigorously). With this in mind, the solution should become clear (hint: use countable additivity).

Using similar reasoning by drawings/illustrations, you should be able to solve your second problem.

(Note: I did not create the illustrations in this post.)

$\endgroup$
3
  • $\begingroup$ I appreciate the graphics! $\endgroup$
    – vladimirm
    Commented Aug 19, 2014 at 10:05
  • 2
    $\begingroup$ I do not understand why is it not the case that $P\left( \bigcup_{i=1}^{\infty}A_i\right )=P\left(A_1 \right )$ because of $ \bigcup_{i=1}^{\infty}A_i=A_1 $. Why do we need the limit in the proof of continuity from above? $\endgroup$
    – Fazzolini
    Commented Jan 17, 2015 at 13:20
  • $\begingroup$ @Fazzolini any answer to question above? $\endgroup$ Commented Oct 30, 2018 at 11:43

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .