Problem 36

Problem: A point is uniformly distributed within the disk of radius 1. That

is, its density is
f (x, y) = C, for x2 + y 2 ≤ 1.
Find the probability that its distance from the origin is less than x, where
0 ≤ x ≤ 1.
Solution: Since the density is uniform, the constant C is determined by the
condition Z
C dx dy = 1.
x2 +y 2 ≤1

The area of the disk of radius 1 is

π(12 ) = π,

so we have
1
C ·π =1 =⇒ . C=
π
The event {D ≤ x} corresponds to the set of all points within a disk of
radius x. Its area is
πx2 .
Thus, the probability that the point is within this smaller disk is

Area of disk of radius x πx2

P {D ≤ x} = = = x2 .
Area of disk of radius 1 π

P {D ≤ x} = x2 , 0 ≤ x ≤ 1.

Problem 37
Problem: Let X1 , X2 , . . . , Xn be independent random variables, each having
a uniform distribution over (0, 1). Define

M = max{X1 , X2 , . . . , Xn }.

Show that the cumulative distribution function (CDF) of M , denoted by FM (x),

is given by
FM (x) = xn , 0 ≤ x ≤ 1.
What is the probability density function (PDF) of M ?
Solution: The event {M ≤ x} means that all of the Xi are less than or
equal to x:
P {M ≤ x} = P {X1 ≤ x, X2 ≤ x, . . . , Xn ≤ x}.

1
Since the Xi are independent and uniformly distributed on (0, 1), we have for
each i,
P {Xi ≤ x} = x, for 0 ≤ x ≤ 1.
Thus,
FM (x) = P {M ≤ x} = xn , 0 ≤ x ≤ 1.

FM (x) = xn , 0 ≤ x ≤ 1.
The probability density function (PDF) of M is the derivative of its CDF:
d d n
fM (x) = FM (x) = (x ) = nxn−1 , 0 ≤ x ≤ 1.
dx dx

fM (x) = nxn−1 , 0 ≤ x ≤ 1.

Problem 50
Problem: Let c be a constant. Show that
(a) Var(cX) = c2 Var(X);
(b) Var(c + X) = Var(X).

Solution:

(a) We start with the definition of the variance:

Var(cX) = E (cX − E[cX])2 .
 

Since E[cX] = c E[X], we can write

Var(cX) = E (cX − c E[X])2 = E c2 (X − E[X])2 .
   

Because c2 is a constant, it factors out of the expectation:

Var(cX) = c2 E (X − E[X])2 = c2 Var(X).
 

(b) Using the definition of variance again, we have:

Var(c + X) = E (c + X − E[c + X])2 .
 

Since E[c + X] = c + E[X], it follows that:

Var(c + X) = E (c + X − (c + E[X]))2 = E (X − E[X])2 = Var(X).
   

Thus, we have shown that:

Var(cX) = c2 Var(X) and Var(c + X) = Var(X).

2
Problem 53
Problem: If X is uniformly distributed over (0, 1), calculate E[X n ] and Var(X n ).

Solution:
For a uniform (0, 1) random variable X, the expectation of X n is given by:
Z 1
n
E[X ] = xn dx.
0

Evaluating the integral:

1
E[X n ] = .
n+1
Next, we calculate E[(X n )2 ]:
Z 1
2n 1
E[X ]= x2n dx = .
0 2n + 1

Using the formula for variance:

Var(X n ) = E[X 2n ] − (E[X n ])2 ,

we get:
 2
1 1
Var(X n ) = − .
2n + 1 n+1
 2
n 1 1 n 1
E[X ] = , Var(X ) = − .
n+1 2n + 1 n+1

Problem 55
Problem: Suppose that the joint probability mass function of X and Y is given
by:
j −2λ λj
 
P (X = i, Y = j) = e , 0 ≤ i ≤ j.
i j!
(a) Find the probability mass function (PMF) of Y . (b) Find the probability
mass function (PMF) of X. (c) Find the probability mass function (PMF) of
Y − X.

Solution:

(a) PMF of Y : To find P (Y = j), sum P (X = i, Y = j) over all possible

values of i:

3
 j  
X j λj
P (Y = j) = e−2λ .
i=0
i j!
Using the binomial theorem:
j  
X j i j−i
11 = 2j ,
i=0
i

we get:
(2λ)j
P (Y = j) = e−2λ .
j!
Thus, Y ∼ Poisson(2λ).

(2λ)j
P (Y = j) = e−2λ .
j!

(b) PMF of X: To find P (X = i), sum P (X = i, Y = j) over all j ≥ i:

∞  
X j λj
P (X = i) = e−2λ .
j=i
i j!

Rewriting the sum in terms of k = j − i:

∞
1 −2λ X λk+i
P (X = i) = e .
i! k!
k=0

Since the remaining sum is the Taylor series for eλ , we obtain:

λi
P (X = i) = e−λ .
i!
Thus, X ∼ Poisson(λ).

λi
P (X = i) = e−λ .
i!

(c) PMF of Y − X: Define k = Y − X. Then,

P (X = i, Y − X = k) = P (X = i, Y = k + i).
From the given joint PMF:

k + i −2λ λk+i
 
P (X = i, Y = k + i) = e .
i (k + i)!
Rewriting:

4
 λi −λ λk
P (X = i, Y − X = k) = e−λe .
i! k!
This shows that X ∼ Poisson(λ) and Y − X ∼ Poisson(λ), and they are
independent.
Thus, the PMF of Y − X is:

λk
P (Y − X = k) = e−λ .
k!

Problem 63
Problem: Calculate the moment generating function (MGF) of a geometric
random variable.

Solution:
Let X be a geometric random variable with probability mass function (PMF):

P (X = n) = (1 − p)n−1 p, n = 1, 2, 3, . . .
The moment generating function (MGF) is given by:
∞
X
ϕ(t) = E[etX ] = etn P (X = n).
n=1

Substituting the PMF:

∞
X
ϕ(t) = etn (1 − p)n−1 p.
n=1

Factoring out constants:

∞
X
n−1
ϕ(t) = pet (1 − p)et .
n=1

The summation is a geometric series with first term 1 and common ratio
(1 − p)et , which converges for (1 − p)et < 1:
∞
X 1
rn = , for |r| < 1.
n=0
1−r
Applying this result:
∞
X
n−1 1
(1 − p)et = .
n=1
1 − (1 − p)et
Thus, the MGF simplifies to:

5
 pet 1
ϕ(t) = , for et < .
1 − (1 − p)et 1−p

pet
ϕ(t) = , for t < − ln(1 − p).
1 − (1 − p)et

Problem 76
Problem: Let X and Y be independent random variables with means µX and
2
µY and variances σX and σY2 . Show that:
2 2
Var(XY ) = σX σY + µ2Y σX
2
+ µ2X σY2 .

Solution:
The expectation of the product is given by:

E[XY ] = E[X]E[Y ] = µX µY .
Next, we compute E[(XY )2 ]:

E[(XY )2 ] = E[X 2 Y 2 ].
Since X and Y are independent, we have:

E[X 2 Y 2 ] = E[X 2 ]E[Y 2 ].

Using the identity:

E[X 2 ] = Var(X) + (E[X])2 = σX

2
+ µ2X ,
and similarly,

E[Y 2 ] = σY2 + µ2Y ,

we obtain:

E[X 2 Y 2 ] = (σX
2
+ µ2X )(σY2 + µ2Y ).
Now, computing the variance:

Var(XY ) = E[(XY )2 ] − (E[XY ])2 .

Substituting the values:
2
Var(XY ) = (σX + µ2X )(σY2 + µ2Y ) − µ2X µ2Y .
Expanding the product:

6
 2 2
Var(XY ) = σX σY + µ2X σY2 + µ2Y σX
2
+ µ2X µ2Y − µ2X µ2Y .
Simplifying:
2 2
Var(XY ) = σX σY + µ2Y σX
2
+ µ2X σY2 .

2 2
Var(XY ) = σX σY + µ2Y σX
2
+ µ2X σY2 .

Problem 79
Problem: With K(t) = log(E[etX ]), show that:

K ′ (0) = E[X], K ′′ (0) = Var(X).

Solution:
The moment generating function (MGF) of X is:

MX (t) = E[etX ].
The cumulant generating function (CGF) is:

K(t) = log MX (t) = log E[etX ].

Differentiating K(t):

E[XetX ]
K ′ (t) = .
E[etX ]
Evaluating at t = 0:

E[Xe0X ] E[X]
K ′ (0) = = = E[X].
E[e0X ] 1
Now, differentiating again:

E[X 2 etX ]E[etX ] − E 2 [XetX ]

K ′′ (t) = .
E 2 [etX ]
Evaluating at t = 0:

E[X 2 ] − E 2 [X]
K ′′ (0) = = E[X 2 ] − E 2 [X] = Var(X).
1
Thus, we conclude:

K ′ (0) = E[X], K ′′ (0) = Var(X).