Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
44 views

Problem Set Conditional Probability

The document provides an overview of conditional probability concepts and examples through solved problems and practice problems. It begins with references for reviewing conditional probability. It then presents 6 solved problems demonstrating properties like conditional probability being a valid probability space and the law of total probability. Finally, it provides 6 practice problems and 1 associated problem on conditional probability concepts like the Monty Hall problem.

Uploaded by

Duppalapudi Sainadh
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
44 views

Problem Set Conditional Probability

The document provides an overview of conditional probability concepts and examples through solved problems and practice problems. It begins with references for reviewing conditional probability. It then presents 6 solved problems demonstrating properties like conditional probability being a valid probability space and the law of total probability. Finally, it provides 6 practice problems and 1 associated problem on conditional probability concepts like the Monty Hall problem.

Uploaded by

Duppalapudi Sainadh
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

EE5110w: Probability Foundations for Electrical Engineers

January - April 2022, Conditional Probability

1 References
• See Chapter 1: Sample Space and Probability, Bertsekas and Tsitsiklis,
for a review of conditional probability.

• See Chapter 3: Conditional Probability and Independence, Sheldon M


Ross, for a review of conditional probability.

2 Solved Problems
1. Consider a probability space (Ω, F, P). Show that the conditional
probability P(·|A), for A ∈ F such that P(A) > 0 inspires a valid
probability space.
We need to show that (Ω, F, P(·|A)) is a valid probability space, where
A ∈ F, P(A) > 0 and P(B|A) for B ∈ F is defined as P(B|A) = P(AB) P(A) .
Clearly, Ω is a valid sample space, and F is a valid event space. In fact,
(Ω, F) remains a valid measurable space independent of the probability
measure assigned.
Hence, we need only to prove that P(·|A) is a valid probability measure
on the measurable space (Ω, F). (We note that P(·|A) is well defined
when P(A) > 0).
To show that P(·|A) is a valid probability measure, we need to validate
that the conditional probability satisfies the axioms of probability.
P(AB)
(a) Non-negativity: P(B|A) = P(A) . As P(AB) ≥ 0 and P(A) > 0,
we have P(B|A) ≥ 0.
(b) Normalized:

P(AΩ) P(A)
P(Ω|A) = = =1
P(A) P(A)

Hence, the conditional probability measure is normalized.


(c) Countably additive: Let B1 , B2 , · · · be a disjoint collection of
events in F.
∞ ∞
! !
[ 1 [
P Bi |A = P ABi
P(A)
i=1 i=1

1
But AB1 , AB2 , · · · are disjoint as well. Hence, using countable addi-
tivity of the probability measure, we have,
∞ ∞
! !
[ 1 [
P Bi |A = P ABi
P(A)
i=1 i=1
∞
1 X
= P(ABi )
P(A)
i=1
∞
X P(ABi )
=
P(A)
i=1
X∞
= P(Bi |A)
i=1

which is the desired result. 


n
 
T
2. Let A1 , A2 , · · · be events in F such that P Ai > 0 for any n.
i=1
Then, show that
∞ ∞
!
\ Y
P Ai = P(Ai |A1 ∩ · · · ∩ Ai−1 )
i=1 i=1

Let A1 , A2 , · · · be events in F. From the continuity theorem, we have,


∞ n
! !
\ \
P Ai = lim P Ai
n→∞
i=1 i=1

Using the multiplicative rule of the conditional probabilities for any


finite n, we have,
∞ n
! !
\ \
P Ai = lim P Ai
n→∞
i=1 i=1
n
Y
= lim P(Ai |A1 · · · Ai−1 )
n→∞
i=1
∞
Y
= P(Ai |A1 · · · Ai−1 )
i=1

where the last equality


Qn identifies the limit of the monotone (bounded)
infinite sequence { i=1 P(Ai |A1 · · · Ai−1 )}. 
3. Let A1 , A2 , · · · be a partition of Ω such that P(Ai ) > 0 for all i. Then,
show that, for any B ∈ F
∞
X
P(B) = P(B|Ai )P(Ai )
i=1

2
The result follows easily from finite additivity.
∞ ∞
! !
[ [
P(B) = P B ∩ Ai = P BAi
i=1 i=1
∞
X
= P(BAi )
i=1
∞
X
= P(Ai )P(B|Ai )
i=1

4. (Ross) A student is taking a one-hour-time-limit makeup examination.


Suppose the probability that the student will finish the exam in less
than x hours is x2 for all 0 ≤ x ≤ 1. Then, given that the student is
still working after 3/4 hour, what is the conditional probability that
the full hour is used?
Let us assume that an appropriate probability space (Ω, F, P) exists.
Let Ax denote the event that the student will finish the exam in less
than x hours. For 0 ≤ x ≤ 1, we know that P(Ax ) = x2 .
Let B denote the event that the student is still working after 0.75
hour. Note that B = Ac0.75 . We need to compute the probability of
the event that the full hour is used for the exam given the knowledge
of occurrence of the event B, i.e., P(Ac1 |B) = 1 − P(A1 |B).

P(BA1 )
P(A1 |B) =
P(B)

We note that BA1 = A1 − A0.75 and P(BA1 ) = P(A1 ) − P(A0.75 )


(since A0.75 ⊂ A1 ). From the description, we have P(A1 ) = 0.5 and
P(A0.75 ) = 83 . Hence,
1 3 1
P(BA1 ) = − =
2 8 8
Also, P(B) = 1 − P(A0.75 ) = 58 . Now,
1
P(BA1 ) 8 1
P(A1 |B) = = 5 =
P(B) 8
5

Then, 1 − P(A1 |B) = 45 . 

3
3 Practice Problems
1. (Bertsekas and Tsitsiklis) We are given three coins: one has heads in
both faces, the second has tails in both faces, and the third has a head
in one face and a tail in the other. We choose a coin at random, toss
it, and the result is heads. What is the probability that the opposite
face is tails?

2. (Bertsekas and Tsitsiklis) A batch of one hundred items is inspected by


testing four randomly selected items. If one of the four is defective, the
batch is rejected. What is the probability that the batch is accepted
if it contains five defectives?

3. State and prove the conditional version of the total probability theo-
rem.

4. (Grimmett and Stirzaker) Assume that P(A)P(B) > 0. Show that if


P(A|B) > P(A), then P(B|A) > P(B).

5. (Grimmett and Stirzaker) There are n urns of which the rth contains
r −1 red balls and n−r magenta balls. You pick an urn at random and
remove two balls at random without replacement. Find the probability
that:

(a) the second ball is magenta;


(b) the second ball is magenta, given that the first is magenta.

6. (Ross) Show that if P(A) > 0, then P(AB|A) ≥ P(AB|A ∪ B).

3.1 Associated Problems


1. Read about Monty Hall Problem (see Wikipedia for a detailed discus-
sion). Can you write a tree representation for the Monty Hall problem
identifying the probabilities along the path.

2. Read about Bertrand’s paradox (see Wikipedia for a detailed discus-


sion).

You might also like