Scribd
Upload
163 views2 pages

Bayes' Theorem

1. The document introduces the average rule and Bayes' rule for probability. The average rule states that the probability of an event A is equal to the sum of the conditional probabilities of A given a partition of events, weighted by the probabilities of those partition events. 2. Bayes' rule provides a formula for calculating the conditional probability of an event given another event, based on the prior probabilities and conditional probabilities in both directions between the two events. 3. Examples of interpreting and applying the average rule and Bayes' rule to probability problems are presented, including recognizing problems involving conditional probabilities and Bayes' rule.

Uploaded by

guipingxie
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
163 views2 pages

Bayes' Theorem

1. The document introduces the average rule and Bayes' rule for probability. The average rule states that the probability of an event A is equal to the sum of the conditional probabilities of A given a partition of events, weighted by the probabilities of those partition events. 2. Bayes' rule provides a formula for calculating the conditional probability of an event given another event, based on the prior probabilities and conditional probabilities in both directions between the two events. 3. Examples of interpreting and applying the average rule and Bayes' rule to probability problems are presented, including recognizing problems involving conditional probabilities and Bayes' rule.

Uploaded by

guipingxie
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Math 461

Introduction to Probability

A.J. Hildebrand

The Average Rule and Bayes Rule


The formulas
1. Partitions: A collection of sets B1 , B2 , . . . , Bn is said to partition the sample space
S if the sets (i) are mutually disjoint and (ii) have as union the entire sample space.
In particular, any set B, together with its complement B c forms a partition of S.
More generally, given a set B, B1 , B2 , . . . , Bn are said to partition a set B if the sets
(i) are mutually disjoint and (ii) their union is B.

2. Average and Bayes Rules, special case: For any events A and B,
(1)
(2)

P (A) = P (A|B)P (B) + P (A|B c )P (B c ) (Average Rule)


P (A|B)P (B)
(Bayes Rule)
P (B|A) =
P (A|B)P (B) + P (A|B c )P (B c )

3. Average and Bayes Rules, general case: Suppose B1 , B2 , . . . , Bn partition the


sample space. Then for each i = 1, 2, . . . , n and any set A,
(3)
(4)

P (A) = P (A|B1 )P (B1 ) + + P (A|Bn )P (Bn ) (Average Rule)


P (A|Bi )P (Bi )
(Bayes Rule)
P (Bi |A) =
P (A|B1 )P (B1 ) + + P (A|Bn )P (Bn )

Notes
1. Technical note: The probabilities P (A|Bi ) are only defined if P (Bi ) > 0; if P (Bi ) =
0, the corresponding term P (A|Bi )P (Bi ) in the above formulas should be interpreted
as 0 (i.e., it doesnt affect the formulas).
2. Memorizing Bayes Rule: The Average Rule says that the expression appearing
in the denominator in Bayes Rule is equal to P (A). Thus, Bayes rule could have
also been stated in the form
(5)

P (Bi |A) =

P (A|Bi )P (Bi )
.
P (A)

You could get by memorizing the average rule (3) along with the above simpler version
of Bayes Rule (5). However, I recommend memorizing Bayes Rule in the full form
(4), since that is the form that you normally need in applications.
3. General versus special case of Bayes Rule: Many, but not all, applications of
Bayes Rule involve only the special case when the simpler formula (2) can be used.
However, for more general problems, one does need the more complicated formula
(4). I recommend to memorize both formulas.

Math 461

Introduction to Probability

A.J. Hildebrand

Interpretations and Applications


1. Interpretation of the Average Rule: A partition B1 , . . . , Bn of S can be interpreted as a set of mutually exclusive conditions that exhaust all possibilities. With
this interpretation, the average rule says that the (overall) probability of A is equal
to a weighted average of all the conditional probabilities, P (A|Bi ), of A with respect
to Bi , with the weigths being the probabilities with which these conditions occur.
2. Interpretation of Bayes Rule: Bayes Rule can be interpreted in terms of prior
and posterior probabilities. The prior probabilities are P (Bi ), i.e., the (ordinary)
probability that the event Bi occurs. Bayes Rule shows how these probabilities
change if we know that event A has occurred; namely it gives a formula for P (Bi |A),
the conditional probability that Bi occurs given that A has occurred. The latter
probabilities are called posterior probabilities. (The terms prior and posterior
come from Latin and mean before and after.)
3. Recognizing Bayes Rule problems: Bayes Rule is a formula for reversing the
order in conditional probabilities. Many (but not all) conditional probability problems
are of this type. If the probability sought in the problem is a conditional probability
and the same conditional probability, but with the order of events reversed, is given
(or can easily be deduced from the given information), the problem is likely a Bayes
Rule problem. Example: In the blood test problem, the probability sought is that
of someone having the disease given that he/she tests positive, whereas the reverse
conditional probability, that someone tests positive given that he/she has the disease,
is given.
4. Recognizing conditional probabilities: Conditional probabilities are often indicated by words/phrases like given that, or if, or by words implying a subpopulation. Here are some examples of statements (mostly taken from actuarial exam
problems) that refer to a conditional probability, along with their translation into
mathematical language. The give-away words that indicate that a conditional probability is involved are set in italics.
Among independents, McCain leads Bush 54% to 36%.
Translation: P (M |I) = 0.54 and P (B|I) = 0.36, where M =McCain voter,
B=Bush voter, I=Independent.
5 percent of those having the disease test negative.
Translation: P (N |D) = 0.05, where N =test negative, D=has disease.
For each smoker, the probability of dying during the year is 0.05
Translation: P (D|S) = 0.05, where D=dies during the year, S=is a smoker.
A blood test indicates the presence of a disease 95% of the time the disease is
actually present.
Translation: P (P |D) = 0.95, where P =tests positive, D=has disease.

You might also like