blob: 03d76da6be35ae2c303137268eb3056d5ecfde8f (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
|
A probability space is Omega, an event space, and P: E -> [0,+\infty)
Why do we use infinite sums (A_1...A_\infty) instead of just finite sum?
Ex. relevant question: I have a fair coin and I flip it until I get a
tails. Are you sure you'll eventually get tails?
Yes.
Lemma: What are the odds that n coin flips result in no tails?
Complement: What are the odds that n coin flips result in only heads?
Each flip is independent and fair, so:
P(A_1 \cap A_2 \cap \ldots \cap A_n) = 1/2^n
P("""^C) = 1 - 1/2^n
Using an infinite union A_1...A_\omega = lim_n->\infty P(A_1\cap...A_n)
This axiom defines this relation.
Properties of probabilities (can be derived from original axioms):
For all A \in E, P(A^C) = 1-P(A)
For all A,B \in E, if A \subset B, P(A) \leq P(B)
For all A,B \in E, P(A \cup B) = P(A)+P(B)-P(A\cap B)
Can construct Poincare lemma or inclusion/exclusion.
|
