Vous êtes sur la page 1sur 8

Page 1 of 8

Indian Institute of Management Bangalore


PGP 20010-11
Quantitative Methods I
Mid Term Examination (For Sections B and E only)


Maximum Time: 2 hours 30 minutes

Name:____________________________

Maximum Marks: 42 Roll. No.________________ Section____



Question No. 1 2 3 4 5 6 7 Total
Maximum Marks 4 6 4 6 6 2*5 =10 2+2+2 = 6 50

Instructions

This is an open-book (1 Textbook), open-note exam; however, you are NOT allowed to
share materials with other students. Use of calculator is permitted, but NOT computer
(laptop). Please do NOT seek any clarifications.

To get any credit, you must

Circle/clearly indicate your final answer (in the space, whenever provided)

Answer all questions in the Answer Book provided

State any assumptions that you make. Assumptions made should be reasonable

Show calculations and provide reasons to support your answers.

This question paper has 3 pages, including this page.
Page 2 of 8

1. If X

1
and X

2
are the means of two independent samples of size n from a
population with mean and varianceo
2
, find a value for n so that

|X

1
-X

2
| <
o

= .99

Answer

By Central Limit Theorem (CLT)

|X

1
] = p, or|X

1
] =
o
2
n
on |X

2
] = p, or|X

2
] =
o
2
n


Therefore, as the two samples are independent (hence X

1
and X

2
are independent),
we have
|X

1
- X

2
] = , on or|X

1
- X

2
] =
o
2
n

Now,
.99 = |X

1
- X

2
| <
o

= -
o

< X

1
- X

2
<
o

= _
-
o

Vo Vn /
<
X

1
-X

2
Vo Vn /
<
o

Vo Vn /
_ = -

_
n

< <

_
n

= >

_
n

= . =

_
n

= .8 = n = - - .8
2
= .8 =


2. A contractor has found through experience that the lowest bids for a job
(excluding his own bid) is a random variable uniformly distributed over the
interval [3/4C, 2C], where C is the contractor's cost estimate (no profit or loss) of
the job. If the profit is defined as 0, if the contractor does not get the job (his bid
is greater than the lowest bid) and as the difference between his bid and his
estimated cost C, if he gets the job, what should he bid (in terms of C) in order to
maximize his expected profit?

Answer

Let X = Lowest bid for a job; then X ~
3C
4
, . So, the density f(x) of X is

() =

, or



Let kC = the bid of the contractor; then the contractors profit P(x, k) is given by

Page 3 of 8

(, ) = < ,
= - ,

Therefore, the contractors expected profit is given by

|(, )] = - rob| < ] + ( - ) - rob| ]
= ( - ) - ( - ) -

( - )( -)

To find k so that |(, )] is maximum. To maximize |(, )],

o(, )
o
= =

( - + ) = = =



o
2
(, )
o
2
= -

- <

So, the contractor should bid =
3C
2
to maximize his expected profit.


3. Assume that the number of fatal car accidents in a certain state obeys a Poisson
distribution with an average of one per day. What is the probability of more than
10 such accidents in a week? (Make suitable assumption if required.)

Answer

Let X
i
= the number of fatal car accidents on the i
th
day of the week

X
i
= Poisson (
i
= 1). We assume that X
i
s are independent. Then Y, the number of
fatal car accidents in a week is

= X


7
=1
osson _z

7
=1
= _
We need

rob| > ] =
-7

=11

!





4. The lifetime of light bulbs is known to be exponentially distributed with a mean
of 100 hours. What is the probability that at least two out of ten bulbs picked at
random will have lifetime of more than 200 hours?
Page 4 of 8

Answer

Let X = lifetime of a light bulb. It is given that

X ~ p(z =
1
100
)

So,

probability that a bulb picked at random will have lifetime of more than 200 hours

p = |X > ] = _ z
-xx

200
= _ .
-0.01x

200
=
-0.01-200
=
-2
= .

Now, let Y = the number of bulbs with lifetime of more than 200 hours out of the
10 randomly selected bulbs. Then

~ n(n = , p = .)

We want

| ] = - | = ] - | = ] = - . - . = .


5. Assume that family incomes are normally distributed with mean = Rs. 16, 000,
and SD = Rs. 2, 000. Forty percent of the families with income above Rs. 18,000
have exactly two children; fifty percent of the families with income between Rs.
15, 000 and Rs. 18,000 have exactly two children, while sixty percent of the
families with income below Rs. 15,000 have exactly two children. What is the
percentage of families with exactly two children have income below Rs. 15, 000?

Answer

Let X = family income; then X ~ (p = , o = ).

Now,

|X > 8] = >
(8 - )

= | > ] = .8

|X < ] = <
( -)

= | < -.] = .8

| < X < 8] = - |X > 8] - |X < ] = .8
Page 5 of 8

Let A be the event that a family has exactly two children.

Then

||X > 8] = .

|| < X < 8] = .

||X < ] = .

We need (using Bayes Theorem)

|X < |] =
|{X < ]]
|]
=
||X < ]|X < ]
||X < ]|X < ] +|| < X < 8]| < X < 8] + ||X > 8]|X > 8]
=
.8 - .
.8 - . +.8 - . + .8 - .
=
.8
.
= .9


35.95% of families with exactly two children have income below Rs. 15, 000.

6. Suppose that (X, Y) has probability density function

f(x, y) = k(x + y) for 0 x y 1
.
a. Find the value of k.
b. Find the probability density function of X.
c. Find the conditional probability density function of Y given X.
d. Are X and Y independent?
e. Compute E[Y|X].




Answer

a. = ] ] ( + )

0
1
0


= _ {_ + _ ]

0
=

0
1
0
_ {
1
0

+
2
] =

_
2
1
0
=



= =


b. Probability density function of X is

Page 6 of 8

X
() = _ ( +) = {_ + _ ] =
1
0
1
0
1
0
_( - ) +

( - )_
= _ +

] , or

In order for
X
() to be a density function

_
X
() =
1
0

So,

X
() = +

, or

Similarly,

() = +

, or

c. Conditional probability density function of Y given X is

|X
(|) =
(, )

X
()
=
( + )
( +

)
=
( + )
( + )
, or

Again, in order for
|X
(|) to be a density function

= _
|X
(|)
1
x

It can be shown that

_
( + )
( + )
= ( - )( + )(
1
x
+ )

So,

|X
(|) =
( + )
( - )( + )
, or


d. As (, ) =
X
()

() , X and Y are NOT independent.



Page 7 of 8

e. ||X] = ]
|X
1
x
(|) = ]
1
x
2(x+)
(1-x)(3x+1)
=
2
(1-x)(3x+1)
| ] +
1
x
]
2
] =
1
x
2
(1-x)(3x+1)
|
(1-x
2
)
2
+
(1-x
3
)
3
] . (You need not simplify this)



7. Customers of a specific product belong to an age group with mean = 30 years and
a variance = 36 years
2
. 30% of these customers have annual income below Rs. 5
lakhs .

a. What conclusions can the marketing office derive about the percentage of
customers whose age is in the interval (22, 38)?

Here,
p = , o
2
=

Using Chebysheves inequality

||X - p| o] -

2


||X -p| o] = |p - o X p + o] = p - o = = =
8



So,

||X - p| o] -

2
=

= .



b. What can you say about the percentage of customers with annual income
above Rs. 5 lakhs or with age in the interval (22, 38) or both?

Let A be the event that a randomly selected customer has annual income
below Rs. 5 lakhs; then

|] = .

Let B be the event a customer has age between 22 and 38; then

|] = ., on |
C
] = . (assuming equality hold for
Chebyshevs inequality)



Page 8 of 8


We need

|] |] + |] = ., sn oo
i
s nqo



c. What can you say about the percentage of customers with annual income
below Rs. 5 lakhs and with age outside the interval (22, 38)?

We need

|
C
] |] + |
C
] - = .8 -
= -., sn onrron
i
s nqo

In this case, Bonferronis inequality gives a trivial result, and is not useful.

Best Wishes!

Vous aimerez peut-être aussi