p2 Intro

Probability
January 15, 2013
Debdeep Pati
Introductory Example
Women in a given age group who give birth to their first child relatively late in life (after
30) are at greater risk for eventually developing breast cancer over some time period t than
are women who give birth to their first child early in life (before 20).
1. 2000 women aged 45-54 with no breast cancer
2. 1000 in group A: first child before age of 20
3. 1000 group B: first child after age of 30
4. Followed for a number of years
5. Suppose there are 4 new cases of breast cancer in group A and 5 in group B.
6. Can we make any conclusions based on this result?
Sample space, events and probability

1. Sample space: the set of all possible outcomes of an experiment.
2. Flipping a coin, S = {H, T }
3. Flipping two coins, S = {(H, H), (H, T ), (T, H), (T, T )}
4. Breast cancer, S = {Y, N }; Two women (Y, Y ), (Y, N ), (N, Y ), (N, N ).
5. Cholesterol level, S = [0, )
6. Events: An event is a set of outcomes, E, of the sample space S.
7. Flipping two coins, E = {(H, H), (H, T )} is the event that a head appears on the
first coin.
8. For breast cancer E = {(Y, N ), (N, Y )} is the event that one of them get cancer.
9. For cholesterol level, E = [100, 200] is the event that blood pressure is between 100
and 200.
10. Probability of an event: Relative frequency of the set of outcomes of the event over
an indefinitely large number of trials.
1
Axioms of probability
1. Probability of an event A, P (A) satisfies 0 P (A) 1
2. Probability of the sample space P (S) = 1
3. If A and B are two events that cannot both happen at the same time, then P(A or
B occurs) = P(A) + P(B)
Operations on Events
1. Union: A B, Union of A and B: event that either A or B occurs.
2. A = {Y }, B = {N }, A B = {Y, N }.A = {X < 90}, B = {90 X < 95}, A B =
{X < 95}
3. Intersection: A B
4. Intersection of A and B: event that both A and B occur A = {(H, H), (H, T ), (T, H)}, B =
{(H, T ), (T, H), (T, T )}, A B = {(H, T ), (T, H)}
5. Commutative laws A B = B A, A B = B A
6. Associative laws A (B C) = (A B) C
7. Distributive laws (A B) C = (A C) (B C), (A B) C = (A C) (B C)
8. If A and B have no outcomes in common, then we say A and B are mutually exclusive
and denoted by A B =
and represents the outcomes in S that
9. Complement of a set E is denoted by E c or E
are not in E.
10. E = {(H, H), (H, T ), (T, H)}, E c = {(T, T )}.
11. E is contained in F, denoted by E F , meaning all of the outcomes in E are also in
F. F = {(H, H), (H, T ), (T, H)}, E = {(H, T ), (T, H)}.
Venn diagram
Venn diagram is a graphical representation for illustrating logical relations among events
i.e., union, intersection, complement, containment.
Conditional probability
1. A small business with 120 workers and 2 employee benefits, A and B. (Venn diagram
in Figure )
B
Bc
Ac
34
15
49
A
64
7
71
98
22
120
2. Calculate the prob. that an employee chosen at random has the retirement benefit
(A) given that she takes the health-care benefit (B). Restrict our attention to the rst
row: 64/98 If the sample space S is finite, and all outcomes are equally likely, it is
often convenient to compute P (E | F ) by using F as the sample space as
P (E | F ) = |E F |/|F |
Count |E F | in the reduced sample space F. (Note that E F is an event in F since
E F F .)
3. While Biopsy is accurate for breast cancer diagnosis, but expensive and invasive.
An alternative is to use mammogram which is noninvasive and less expensive. The
idea is to have mammogram first and then decide whether to take biopsy based
on mammogram result. A = {mammogram+}, B = {breast cancer} Conditional
3
probability of B given A: P (B | A) = P (A B)/P (A). Relative risk of B given A:

P (B | A)/P (B | Ac ). ( Youll see later that If two events A and B are independent,
then the relative risk is 1).
Suppose that among 100,000 women with negative mammogram 20 will have breast
cancer diagnosed within 2 years, or P (B | Ac ) = 20/105 = 0.0002, whereas 1 woman
in 10 with positive mammograms will have breast cancer diagnosed within 2 years,
or P (B | A) = 0.1. The two events A and B are highly dependent, since P (B |
A)/P (B|Ac ) = .1/.0002 = 500.
4. Total Probability Rule: For any events A and B, P (B) = P (B | A)P (A) + P (B |
Ac )P (Ac )
Example. Cancer A: mammogram positive
B: developing breast cancer in next 2 years
Suppose that 7% of the general population of women will have a positive mammogram. What is the probability of developing breast cancer over the next 2 years
among women in the general population?
P(breast cancer | mammogram+) = .1
P(breast cancer | mammogram-) = .0002
P(B) = P(breast cancer)
= P(breast cancer | mammogram+)P(mammogram+) + P(breast cancer | mammogram)P(mammogram-) = .1(.07) + .0002(.93) = 0.00719
5. Digitalis therapy is often beneficial to patient who have suffered congestive heart
failure, but there is the risk of digitalis intoxication, a serious side effect that is,
moreover, difficult to diagnose. To improve the chances of a correct diagnosis, the
concentration of digitalis in the blood can be measured. Bellar (1971) conducted
a study of the relation of the concentration of digitalis in the blood to digitalis
intoxication in 135 patients. Their results are simplified slightly in the following
table.
Table 1: default
T+
TTotal
D+
25
18
43
D14
78
92
Total
39
96
135
T+ = high blood concentration (positive test) T- = low blood concentration (negative

test) D+ = toxicity (disease present) D- = no toxicity (disease absent) 25 of the 135
patients had a high blood concentration of digitalis and suffered toxicity.
4
Table 2: Digitalis therapy example
T+
TTotal
D+
0.185
0.133
0.318
D0.104
0.578
0.682
Total
0.289
0.711
1
P (T +) = .289, P (D+) = .318 If the patient has high blood concentration (T+),
what is the probability of disease (D+)? P (D+ | T +) = 25/39 = .64 P (D+ | T +) =
P (D + T +)/P (T +) = .185/.289 = .64
Independence
Two events, E and F , are independent if knowing that one had occurred gave us no
information about whether the other had or had not occurred. If
P (E) = P (E | F ), P (F ) = P (F | E)
then E is independent on F. If P (E) 6= P (E | F ), then E is dependent on F.
Definition 1 Two events E and F are independent if P (E F ) = P (E)P (F ). Two events
E and F that are not independent are said to be dependent.
1. A card is randomly selected from a deck. Let E denote the event that the card is an
ace and F the event that it is a spade. Are E and F independent? Knowing that
the card is an ace gives no information about its suit. P (E) = 4/52, P (F ) = 13/52,
and P (E F ) = 1/52 P (E F ) = P (E)P (F )
2. Suppose we are conducting a hypertension-screening program in the home. Consider
all possible diastolic blood-pressure measurements from a mother and her first-born
child. Let A = {mothers DBP > 95}, B = {first-born childs DBP > 80} Suppose
P (A B) = 0.05, P (A) = .1, P (B) = .2 Then P (A B) = .05 > P (A) P (B) = .02,
and the events A, B would be dependent.
3. Multiplication law of probability: If A1 , A2 , . . . , Ak are mutually independent events,
then
P (A1 A2 . . . Ak ) = P (A1 )P (A2 ) P (Ak )
4. Addition law of probability: In the previous example, what is P (A B), the probability that either mothers DBP > 95 or the first born childs DBP > 95? P (AB) =
P (A) + P (B) P (A B) = .1 + .2 .05 = .25

p2 Intro

Uploaded by

Copyright:

Available Formats

p2 Intro

Uploaded by

Document Information

Original Description:

Original Title

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

p2 Intro

Uploaded by

Copyright:

Available Formats

Probability

January 15, 2013

Sample space, events and probability

probability of B given A: P (B | A) = P (A B)/P (A). Relative risk of B given A:

T+ = high blood concentration (positive test) T- = low blood concentration (negative

Table 2: Digitalis therapy example

You might also like

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