MA2040: Probability, Statistics and Stochastic Processes

Problem Set-III
Sivaram Ambikasaran
March 20, 2019

1. If X1 , X2 , . . . , Xn are independent random variables having the same probability density function
fX (x), what is the probability density function for the random variable Y = min{X1 , X2 , . . . , Xn }?
Solution:

P (Y ≥ y) = P (X1 ≥ y, X2 ≥ y, . . . , Xn ≥ y) = P (X1 ≥ y)P (X2 ≥ y) . . . P (Xn ≥ y)

Hence, we obtain that

n
1 − FY (y) = (1 − FX (y))
n−1
fY (y) = nfX (y) (1 − FX (y))

2. A random variable X has a probability density function as shown below.

fX (x)

(0, h) (2, h)

(0, 0)
x
(1, 0) (2, 0)

(a) Determine h
Solution: Area equals one implies h = 1
(b) Determine the cumulative distribution function

0
 if x ≤ 0
Solution: FX (x) = x2 /4 if x ∈ [0, 2]

1 if x ≥ 2


(c) Compute the mean

Solution: We have Z 2 Z 2
x
E (X) = xfX (x)dx = x· dx = 4/3
0 0 2

1
 (d) Compute the variance

Solution: Z 2 Z 2
x
E X2 = x2 fX (x)dx = x2 ·


dx = 2 =⇒ Var (X) = 2/9
0 0 2
(e) Determine the probability that X ∈ (1, 2).

Solution: Z 2 Z 2
P (X ∈ (1, 2)) = fX (x)dx = x/2dx = 3/4
1 1

3. The median m of a probability density function is defined as the value of m such that
Z m Z ∞
f (x)dx = f (x)dx = 1/2
−∞ m

Essentially, the median splits the distribution into two equal halves. Prove that the median is the
best predictor if one wants to minimize the expected value of the absolute error, i.e., E (|X − c|) is
minimized when c is the median of the underlying distribution.
Solution: We have
Z c Z ∞
E (|X − c|) = (c − x) fX (x)dx + (x − c) fX (x)dx
−∞ c

Differentiating with respect to c, we obtain

Z c Z ∞
fX (x)dx = fX (x)dx
−∞ c

We also have that Z c Z ∞

fX (x)dx + fX (x)dx = 1
−∞ c
Hence, this gives us that Z c Z ∞
fX (x)dx = fX (x)dx = 1/2
−∞ c

4. Let X be a random variable, whose pdf is given by

(
0 if x ≤ 0
fX (x) = 2
xe−x /2 if x > 0

Find the pdf for the random variable Y = X 2 .

Solution: Since Y = X 2 is an increasing function on [0, ∞), we have
√
√ dx fX ( y)
fY (y) = fX ( y) = √ = 1/2e−y/2
dy 2 y

5. Let X be a uniform random variable on the interval [0, 1]. Consider the random variable Y = g (X),
where (
1 if x ≤ 1/3
g(x) =
2 else
Find the probability mass function of Y and compute its expected value.

Solution:
P (Y = 1) = P (X ≤ 1/3) = 1/3
P (Y = 2) = P (X ≥ 1/3) = 2/3

2
6. Show the expected value of a random variable X can also be obtained as
Z ∞ Z ∞
E (X) = P (X > x) dx − P (X < −x) dx
0 0

Solution: We have X = X + − X − , where X + = max{X, 0} and X − = max{−X, 0}. Note that

Z ∞
X+ = I (X > t) dt
0

and Z 0
X− = I (X ≤ t) dt
−∞

where I is the indicator function that takes the value 1, when the argument is true and takes the value
0, when the argument is false. We have

E (X) = E X + − X −

which gives us what we want. Note that expectation of the indicator function is nothing but the
probability of the argument.
7. A defective coin minting machine produces coins whose probability of heads is a random variable Y
with PDF (
y exp (y) if y ∈ [0, 1]
fY (y) =
0 otherwise

8. Let the random variables X and Y have a joint PDF, which is uniform over the triangles with vertices
(0, 0), (0, 1) and (1, 0).
(a) Find the joint PDF of X and Y .

Solution: fXY (x, y) = 2 over the triangle.

(b) Find the marginal PDFs.

Solution: Z 1−x
fX (x) = 2dy = 2 (1 − x)
y=0
Z 1−y
fY (y) = 2dx = 2 (1 − y)
x=0

(c) Find the conditional PDFs.

Solution:
1
fX|Y =y =
1−y
1
fY |X=x =
1−x
9. Chennai’s temperature is modeled as a normal random variable with a mean temperature of 34◦ C and
a standard deviation of 5◦ C. What is the probability that the temperature at a randomly chosen time
will exceed 45◦ C?
45 − 34
Solution: z = = 2.2. Φ (2.2) = 0.9861. Hence, desired probability is 1 − 0.9861 = 0.0139.
5

3
10. A surface is ruled with parallel lines, which are at a distance d from each other. Suppose that we throw
a needle of length l on the surface at random. What is the probability that the needle with intersect
one of the lines? (NOTE: You will need to treat the case d < l and d > l separately.)

Solution: Page 161. Bertsekas.

2l
d>l:
πd
 s 
   2
2 d 2l  d 
d < l : arccos + 1− 1−
π l πd l

11. Consider two continuous random variables Y and Z and a random variable X that is equal to Y with a
probability p and equals Z with a probability 1−p. Obtain the pdf of X interms of the pdf’s of Y and Z.

Solution:
FX (x) = P (X ≤ x) = pP (Y ≤ x) + (1 − p) P (Z ≤ x)
Hence,
fX (x) = pfY (x) + (1 − p) fZ (x)