Probability Theorem Proofs

AXIOMATIC PROBABILITY AND POINT SETS
The axioms of Kolmogorov. Let S denote an event set with a probability measure P
defined over it, such that probability of any event A S is given by P (A). Then, the
probability measure obeys the following axioms:
(1) P (A) 0,
(2) P (S) = 1,
(3) If {A1 , A2 , . . . Aj , . . .} is a sequence of mutually exclusive events such that Ai Aj =
for all i, j, then P (A1 A2 Aj ) = P (A1 ) + P (A2 ) + + P (Aj ) + .
The axioms are supplemented by two definitions:
(4) The conditional probability of A given B is defined by
P (A|B) =
P (A B)
,
P (B)
(5) The events A, B are said to be statistically independent if

P (A B) = P (A)P (B).
This set of axioms was provided by Kolmogorov in 1936.
1
The rules of Boolean Algebra. The binary operations of union and intersection are
roughly analogous, respectively, to the arithmetic operations of addition + and multiplication
, and they obey a similar set of laws which have the status of axioms:
Commutative law:
Associative law:
Distributive law:
Idempotency law:
A B = B A,
A B = B A,
(A B) C = A (B C),
(A B) C = A (B C),
A (B C) = (A B) (A C),
A (B C) = (A B) (A C),
A A = A,
A A = A.
De Morgans Rules concerning complementation are also essential:

(A B)c = Ac B c
and
(A B)c = Ac B c .
Amongst other useful results are those concerning the null set and the universl set S:
(i) A Ac = S,
(ii) A Ac = ,
(iii) A S = S,
(iv) A S = A,
(v) A = A,
(vi) A = .
2
LEMMA: the probability of the complementary event. If A and Ac are complementary events, then
P (Ac ) = 1 P (A).
Proof. There are
Therefore, by Axiom 3,
A Ac = S
and
A Ac = ,
P (A Ac ) = P (A) + P (Ac ) = 1,
since P (A Ac ) = P (S) = 1, whence P (Ac ) = 1 P (A).
LEMMA: the probability of the null event. The probability of the null event is
P () = 0.
Proof. Axiom 3 implies that
P (S ) = P (S) + P (),
since S and are disjoint sets by definition, i.e. S = . But also S = S, so
P (S ) = P (S) = 1,
where the second equality is from Axiom 2. Therefore,
P (S ) = P (S) + P () = P (S) = 1,
so P () = 0.
3
THEOREM: independence and the complementary event. If A, B are statistically

independent such that P (A B) = P (A)P (B), then A, B c are also statistically independent
such that P (A B c ) = P (A)P (B c ).
Proof. Consider
A = A (B B c ) = (A B) (A B c ).
The final expression denotes the union of disjoint sets, so there is
P (A) = P (A B) + P (A B c ).
Since, by assumption, there is P (A B) = P (A)P (B), it follows that
P (A B c ) = P (A) P (A B) = P (A) P (A)P (B)
= P (A){1 P (B)} = P (A)P (B c ).
THEOREM: the union of of events. The probability that either A or B will happen or
that both will happen is the probability of A happening plus the probability of B happening
less the probability of the joint occurrence of A and B:
P (A B) = P (A) + P (B) P (A B)
Proof. There is A (B Ac ) = (A B) (A Ac ) = A B, which is to say that A B can
be expressed as the union of two disjoint sets. Therefore, according to axiom 3, there is
P (A B) = P (A) + P (B Ac ).
But B = B (A Ac ) = (B A) (B Ac ) is also the union of two disjoint sets, so there
is also
P (B) = P (B A) + P (B Ac ) = P (B Ac ) = P (B) P (B A).
Substituting the latter expression into the one above gives
P (A B) = P (A) + P (B) P (A B).
BAYES THEOREM
Observe that the formula for conditional probability implies that
whence we
P (A B) = P (A|B)P (B) = P (B|A)P (A),

P (A|B) =
P (B|A)P (A)
.
P (B)
Consider a set = {H1 , H2 , . . . , Hn }, wherein Hi Hj = ; i 6= j, which comprises all possible

explanations of an event E. The evidence of E will cast some light upon the likelihoods of
the hypotheses. The posterior likelihood of an hypothesis Hi in the light of the event E is
P (E|Hi )P (Hi )
, where
P (E)
X
X
P (E) =
P (E Hi ) =
P (E|Hi )P (Hi ).
P (Hi |E) =
We use the following terminology:

P (Hi ) is the prior likelihood of the ith hypothesis Hi ,
P (Hi |E) is the posterior likelihood of the ith hypothesis,
P (E|Hi ) is the conditional probability of the event E under the hypothesis Hi ,

P (E) is the unconditional probability of the event E.
6
A Bayesian Problem. The dawn train collects any milk churns that were left on the
platform of Worplesham station on weekdays. Churns are left on three of the five days. A
man arrives at the station not knowing the exact time and thinking that there is a fifty-fifty
chance that he has missed the train. Then he notices that there are no milk churns on the
platform. How should he reassess the chances that he has missed the train?
Answer: Let T be the event that the train has already passed and let N be the event of
there being no milk churns on the platform when I arrive. We have
P (T |N ) =
P (N |T )P (T )
P (N )
with
P (N ) = P (N |T )P (T ) + P (N |T c )P (T c ).
We take
P (N |T ) = 1,
Then
P (T ) = P (T c ) =
1
2
and P (N |T c ) =
1 2 1
7
P (N ) = 1. + . =
2 5 2
10
1 10
5
and P (T |N ) = 1. .
= .
2 7
7
2
.
5

Probability Theorem Proofs

Uploaded by

Copyright:

Available Formats

Probability Theorem Proofs

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Probability Theorem Proofs

Uploaded by

Copyright:

Available Formats

AXIOMATIC PROBABILITY AND POINT SETS

(5) The events A, B are said to be statistically independent if

De Morgans Rules concerning complementation are also essential:

since P (A Ac ) = P (S) = 1, whence P (Ac ) = 1 P (A).

THEOREM: independence and the complementary event. If A, B are statistically

P (A B) = P (A|B)P (B) = P (B|A)P (A),

Consider a set = {H1 , H2 , . . . , Hn }, wherein Hi Hj = ; i 6= j, which comprises all possible

We use the following terminology:

P (E|Hi ) is the conditional probability of the event E under the hypothesis Hi ,

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.