Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)

Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
MSc DS (CMI), MSQMS (ISI), Mtech CS & CrS (ISI)
Assignment

66
Topic - CONTINUOUS RANDOM VARIABLES
(Statistics & Probability)

80
1 USEFUL DEFINITION AND RESULTS ——————————–
1.1 BASIC DEFINITIONS
1. Sample space : Sample space is the set of all possible outcomes in an experiment. Sample space is
denoted by S.

81
2. Event : Any subset E of a sample space S is called an Event.
3. Probability : Probability is a function P : Ω → [0, 1] ,
where Ω = P(S) (power set of S) and S=Sample space,
such that
(1) 0 ≤ P (E) ≤ 1 , ∀E ⊆ S
(2) P (S) = 1
0
(3) For any set of events E1 , E2 , E3 , . . . such that Ei ∩ Ej = ϕ when i ̸= j, then
∞ ∞
!
[ X
P En = P (En ).
79
n=1 n=1

And P (E) is called the probability of the event E.

4. Random variable : Random variable is a function X : S → R , where S=Sample space.
5. Cumulative distribution function (c.d.f.) : Cumulative distribution function (c.d.f.) or Distribution
function (d.f.) of a random variable X is a function F : R → [0, 1] defined by F (x) = P (X ≤ x).
i(

6. Continuous random variable : We say a random variable X is a continuous random variable if its
c.d.f. F is continuous at every point X = x .
⋆ ⋆ From now in this chapter, "Random Variable" is written to mean "Continuous Random Variable".
7. Probability density function (p.d.f.): The probability density function (when exists) of a continuous
Rx
random variable X with c.d.f. F , is a non-negative function f such that F (x) = −∞ f (u), for −∞ < x <
∞. [Here this is Reimann Integration]
pa

For a continuous random variable X, p.d.f. may not be unique but c.d.f. must be unique.
Note that, when F (x) is differentiable at x, the p.d.f. of X at x, f (x) = d
dx F (x) = F ′ (x).
8. Joint cumulative distribution function (c.d.f.) : Let X1 , X2 , . . . Xr are r continuous random vari-
ables. Joint cumulative distribution function (Joint c.d.f.) or Joint distribution function (Joint d.f.) of
these r continuous random variables is defined as
F (x1 , x2 , . . . , xr ) := P (X1 ≤ x1 , X2 ≤ x2 , . . . , Xr ≤ xr ).
Pa

Obviously, The joint cumulative distribution function of two random variables X and Y is defined as
F (x, y) := P (X ≤ x, Y ≤ y). Sometimes F (x, y) is denoted as FX,Y (x, y).
If F (x, y) is the joint c.d.f. of X and Y then the marginal c.d.f.’s of X and Y are
F (x) = limy→∞ F (x, y) for any x ∈ R and F (y) = limx→∞ F (x, y) for any y ∈ R.

CAREER FIELD 1
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
Notice that, if F (X, Y ) is the joint c.d.f. of the random variables X and Y then
limx→−∞ F (x, y) = 0 for any y ∈ R,
limy→−∞ F (x, y) = 0 for any x ∈ R,
limx→−∞ limy→−∞ F (x, y) = 0 and
limx→∞ limy→∞ F (x, y) = 1

66
9. Joint probability density function (Joint p.d.f.): Suppose X1 , X2 , . . . Xr are r continuous random
variables with jointRc.d.f.R F (x1 , Rx2 , . . . xr ). If there is a non-negative function f such that
x1 x2 xr
F (x1 , x2 , . . . xr ) = −∞ −∞
. . . −∞ f (u1 , u2 , . . . , ur )du1 du2 . . . dur , then f is called a Joint p.d.f. of the
random variables X1 , X2 , . . . , Xr .
R∞ R∞ R∞
And obviously −∞ −∞ . . . −∞ f (u1 , u2 , . . . , ur )du1 du2 . . . dur = 1.
When F (x1 , x2 , . . . xr ) is differential at (x1 , x2 , . . . , xr ) then
∂r

80
F (x1 , x2 , . . . , xr ) = f (x1 , x2 , . . . , xr ).
∂x1 ∂x2 . . . ∂xr
Suppose f (x, y) is a joint p.d.f. of two continuous random variables X and Y . This f (x, y) can be denoted
as fX,Y (x, y) also.
10. Marginal probability density function : Let X, Y are two continuous random variables with a joint
R∞ R∞
p.d.f. f (x, y) , then the marginal p.d.f.’s are given by fX (x) = −∞ f (x, v)dv and fY (y) = −∞ f (u, y)du.
R∞ R∞
Obviously −∞ −∞ f (x, y) = 1 .
81
11. Independent random variable : Let X1 , X2 , . . . Xr are r continuous random variables having c.d.f.’s
FX1 , FX2 , . . . , FXr respectively. These random variables are said to be mutually independent if their
joint c.d.f. is given by F (x1 , x2 , . . . , xr ) = FX1 (x1 )FX2 (x2 ) . . . FXr (xr ) .
So, two random variables X and Y are said to be independent if P (X ≤ x, Y ≤ y) = P (X ≤ x)P (Y ≤ y).
0
Let X1 , X2 , . . . Xr are r random variables having p.d.f.’s fX1 , fX2 , . . . , fXr respectively. These random
variables will be mutually independent if their joint p.d.f. is given by
f (x1 , x2 , . . . , xr ) = fX1 (x1 )fX2 (x2 ) . . . fXr (xr ).
79

Two random variables X and Y will be independent, if their joint p.d.f. is given by
f (x, y) = fX (x)fY (y). (WherefX (x) and fY (y) are p.d.f.’s of X and Y respectively.)
12. Independent identically distributed random variable (i.i.d.) : Let X1 , X2 , . . . Xr are r mutually
independent continuous random variables having p.d.f.’s fX1 , fX2 , . . . , fXr respectively. If the p.d.f.’s are
exactly same for these r random variables (means, fX1 = fX2 = . . . = fXr ), then we say X1 , X2 , . . . Xr
are i.i.d. random variables or Independent identically distributed random variables.
i(

Example: Let X1 and X2 be two independent random variables with Xi ∼ N (0, 1), for i = 1, 2 . Then
we say X1 and X2 are i.i.d.’s.
13. Conditional p.d.f. : Let the joint p.d.f. of X and Y is fX,Y (x, y). The conditional p.d.f. of X given
Y = y (here y is a fixed value) is defined as

fX,Y (x, y)
fX|Y (x|y) := ,
fY (y)
pa

And we say, fX|Y (x|y) is the conditional p.d.f. of the random variable X/Y =y (means, X given Y = y).
If X and Y are independent random variables, then fX|Y (x|y) = fX (x).
14. Expectation / Mean:
R +∞
Let X be a random variable having the p.d.f. fX (x). If −∞ |x|fX (x)dx < ∞, then we say that X has
fnite expectation. And the expectation or mean ,
Pa

Z +∞
E(X) := xfX (x)dx
−∞

E(X) is denoted as EX, µX or µ.

CAREER FIELD 2
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
□ Result : Let X be a random variable having the p.d.f. fX (x) and g(x) is a real valued function. If
R +∞
−∞
|g(x)|fX (x)dx < ∞, then for Z = g(X) , E(Z) exists. And the expectation or mean of Z or g(X),
Z +∞
E(Z) = E(g(X)) = g(x)fX (x)dx.
−∞

66
R +∞ R +∞
Example : E(X 2 ) = −∞ x2 fX (x)dx , E(X 3 + 2X + 1) = −∞ (x3 + 2x + 1)fX (x)dx , E(eX ) =
R +∞ x R +∞
−∞
e fX (x)dx , E((X − µ)2 ) = −∞ (x − µ)2 fX (x)dx.
□ □ Definition : Let X and Y be two random variables having the joint p.d.f. fX,Y (x, y) and g(x, y) is a
R +∞ R +∞
real valued function. If −∞ −∞ |g(x, y)|fX,Y (x, y)dxdy < ∞, then for Z = g(X, Y ) we say that E(Z)
exists. And the expectation or mean of Z or g(X, Y ),
Z +∞ Z +∞ Z +∞ Z +∞

80
E(Z) = E(g(X, Y )) := g(x, y)fX,Y (x, y)dxdy = g(x, y)fX,Y (x, y)dxdy.
−∞ −∞ −∞ −∞
R +∞ R +∞ R +∞
Example : E(X + Y ) = −∞ −∞ (x + y)fX,Y (x, y)dxdy , E(XY ) = −∞
(xy)fX,Y (x, y)dxdy ,
R +∞ R +∞
E(X 2 + 3Y ) = −∞ −∞ (x2 + 3y)fX,Y (x, y)dxdy .
X1 + X2 + . . . + Xn
15. Sample mean : If X1 , X2 , . . . Xn are i.i.d. random variables , then the random variable X¯n =
is called the sample mean.
16. Variance and Standard deviation :
81
Let X be a random variable with expectation E(X) = µX . The variance of X is defined as
V ar(X) := E(X − µX )2 = E(X − E(X))2 = E(X 2 ) − (E(X))2 .
n

Always, V ar(X) ≥ 0.
0
2
V ar(X) is denoted as σX or σ 2 .
p
The Standard deviation of a random variable X is defined as σ(X) = σX := V ar(X).
79

17. Sample variance : If X1 , X2 , . . . Xn are i.i.d. random variables with sample mean X¯n , then the random
variable
1 Pn
Sn2 = (Xj − X¯n )2 is called the sample variance.
n − 1 j=1
18. Covariance and Correlation coefficient:
Let X and Y be two random variables with expectations E(X) = µX and E(Y ) = µY . The covariance of
X and Y is defined as
i(

Cov(X, Y ) := E{(X − µX )(Y − µY )} = E(XY ) − E(X)E(Y ).

So, obviously, Cov(X, X) = V ar(X) and Cov(X, Y ) = Cov(Y, X).

The Correlation coefficient of two random variables X and Y is defined as
Cov(X, Y )
ρ(X, Y ) = ρX,Y := p .
pa

V ar(X)V ar(Y )

So, obviously, ρ(X, Y ) = ρ(Y, X).

19. Median:
The median of X is the value x for which P (X ≤ x) = 0.5 , i.e. P (X ≤ x) = P (X ≥ x). Equivalently,
the median is the value x satisfying F (x) = 0.5 .
Pa

20. Quantile:
1
The rth quantile of X is the value x such that P (X ≤ x) = .
r
p
The pth percentile of X is the value x such that P (X ≤ x) = . So, rth quantile = 100rth percentile.
100

CAREER FIELD 3
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
R∞
21. Gamma function: If x > 0, then Gamma x or Γ(x) = 0
e−t tx−1 dt .
Properties and results:
i) If x > 0 then Γ(x + 1) = xΓ(x).
ii) Γ(1) = 1.√
iii) Γ( 12 ) = π.
iv) If n ∈ N, then Γ(n + 1) = n! .

66
R1
v) If x > 0, Γ(x) = 0 (− log t)x−1 dt .
R1
22. Beta function: If x > 0 and y > 0, then Beta (x, y) or β(x, y) = 0
tx−1 (1 − t)y−1 dt .
Properties and results:
i) β(x, y) = β(y, x), where x, y > 0.
Γ(x)Γ(y)
ii) β(x, y) = , where x, y > 0.
Γ(x + y)

80
1.2 IMPORTANT DISTRIBUTIONS
1. Uniform Distribution :
We say, the random variable, X ∼ Uniform [a, b] or X ∼ U[a, b] , where −∞ < a < b < ∞
if the p.d.f. of X is 
 1

Note:
fX (x) = b − a
0 81
for a ≤ x ≤ b
otherwise.

We say, the random variable, X ∼ Uniform (a, b) or X ∼ U(a, b) , where −∞ < a < b < ∞
if the p.d.f. of X is 
 1 for a < x < b
fX (x) = b − a
0
0 otherwise.

a+b 1
In both case, E(X) = 2 , V ar(X) = 12 (b − a)2 .
79

2. Gamma Distribution :
We say, the random variable, X ∼ Gamma (α,λ) or X ∼ Γ(α, λ), where α > 0, λ > 0
if the p.d.f. of X is ( α
λ
e−λx xα−1 for x > 0
fX (x) = Γ(α)
0 otherwise.
α α
i(

Here E(X) = λ , V ar(X) = λ2 .

3. Beta Distribution :
We say, the random variable, X ∼ Beta (α,γ), where α > 0, γ > 0
if the p.d.f. of X is
 Γ(α + γ) xα−1 (1 − x)γ−1 for 0 < x < 1


fX (x) = Γ(α)Γ(γ)
0 otherwise.

pa

α αγ
Here E(X) = α+γ , V ar(X) = (α+γ)2 (α+γ+1) .

4. Exponential Distribution :
We say, the random variable, X ∼ Exponential (λ) or X ∼ exp (λ)
if the p.d.f. of X is (
λe−λx for x > 0 (or sometimes x ≥ 0)
fX (x) =
0 otherwise.
Pa

Note that, X ∼ exp (λ) is equivalent to saying X ∼ Gamma (1, λ) .

1 1
Here E(X) = λ , V ar(X) = λ2 .

CAREER FIELD 4
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
5. Normal Distribution / Gaussian Distribution :
We say, the random variable, X ∼ Normal (µ, σ 2 ) or X ∼ N (µ, σ 2 )
if the p.d.f. of X is
1 1 x−µ 2
fX (x) = √ e− 2 ( σ ) for −∞ < x < ∞.
2πσ

66
Here E(X) = µ , V ar(X) = σ 2 .
Standard Normal Distribution :
We say, the random variable, X follows Standard Normal Distribution if X ∼ N (0, 1),
then the p.d.f. of X is
1 x2
fX (x) = √ e− 2 for −∞ < x < ∞.
2π

80
Here E(X) = 0 , V ar(X) = 1 .
Note that:
Suppose X = µ + σZ, then Z ∼ N (0, 1) iff X ∼ N (µ, σ 2 ).

6. Chi-square Distribution or χ2 -distribution :

We say, the random variable, X ∼ χ2 (n) [ χ2 (n) means chi-square distribution with degree of freedom
n, where n ∈ N ]
if the p.d.f. of X is

fX (x) =

Here E(X) = n , V ar(X) = 2n .




0
1 n/2
81
 ( 2 ) e−x/2 xn/2−1 for 0 < x < ∞

Γ( n2 )
otherwise.

Note that, X ∼ χ2 (n) is equivalent to saying X ∼ Gamma ( n2 , 12 ) .

0
7. Cauchy Distribution :
We say, the random variable, X ∼ Cauchy (µ, θ) or X ∼ C(µ, θ), where µ > 0 and θ ∈ R
if the p.d.f. of X is
79
1 1
fX (x) = . for −∞ < x < ∞.
πµ 1 + ( x−θ
µ )
2

Here E(X) = undefined , V ar(X) = undefined .

Standard Cauchy Distribution :
We say, the random variable, X ∼ Cauchy (1, 0),
if the p.d.f. of X is
i(

1
fX (x) = for −∞ < x < ∞.
π(1 + x2 )

Here E(X) = undefined , V ar(X) = undefined .

8. Student-t distribution :
We say, the random variable, X ∼ Student-t (α) or t−distribution with degree of freedom α.
if the p.d.f. of X is
pa

Γ( α+1
2 ) x2 − α+1
fX (x) = √ α (1 + ) 2 for −∞ < x < ∞.
πα.Γ( 2 ) α

α
(
 α−2
 for α > 2
0 for α > 1
Here E(X) = , V ar(X) = ∞ for 1 < α ≤ 2 .
undefined otherwise. 
undefined otherwise.

Pa

9. F-distribution :
We say, the random variable, X ∼ F (k1 , k2 ),
if the p.d.f. of X is

CAREER FIELD 5
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)

k1 +k2 k1 x ( k21 −1)
 k1 . Γ( 2 ) . ( k2 )


for x > 0
fX (x) = k2 Γ( k21 )Γ( k22 ) (1 + k1 x ) k1 +k
2
2
 k2

0 otherwise.

66
2k22 (k1 +k2 −2)
( (
k2
k2 −2 for k2 > 2 k1 (k2 −2)2 (k2 −4) for k2 > 4
Here E(X) = , V ar(X) = .
undefined otherwise. undefined otherwise.

1.3 STANDARD NORMAL DISTRIBUTION TABLE

Let X ∼ N (0, 1), means X is a standard normal random variable, so the p.d.f. of X,

80
1 x2
fX (x) = √ e− 2 for −∞ < x < ∞.
2π
Here E(X) = 0 , V ar(X) = 1 .
The cumulative distribution function of a standard normal random variable X is denoted as Φ(x),
i.e. F (x) = P (X ≤ x) = Φ(x).
So,

Note that,
1
81
Φ(x) = √
1
2π
Z

−∞
x
e−z
2
/2
dz.

(1) fX (0) = √ = 0.3989,

2π
0
(2) fX (x) = fX (−x),
(3) Φ(−x) = 1 − Φ(x).
79

• Here is the graph of p.d.f. of X

i(
pa
Pa

CAREER FIELD 6
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
• Φ(a) is the area under the curve fX (x) upto x = a. See the following picture,

66
• The following is the table of values of Φ(x), which is the standard normal distribution table.

80
0 81
79
i(
pa
Pa

The values of Φ(x) for non-negative x are given in this Table. For negative values of x, Φ(x) can be
obtained from the relation Φ(−x) = 1 − Φ(x), ∀x ∈ R.

CAREER FIELD 7
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
2 USEFUL THEOREMS ——————————————————–
1. Theorem :
Let g be a differentiable strictly monotone (strictly increasing or strictly decreasing) function on an interval
I and g(I) denote the range of g and let g −1 denotes the inverse function of g . Let X be a continuous
random variable having p.d.f. fX (x), such that fX (x) = 0 for x ∈ / I. Then Y = g(X) has p.d.f. fY given

66
d −1 −1
by fY (y) = 0 for y ∈/ g(I) and fY (y) = g (y) fX (g (y)), for y ∈ g(I).
dy
2. Theorem :
Let X, Y be two continuous random variable with joint p.d.f. f . Let Z = X + Y , U = X − Y , V = XY
and W = X/Y . Then the p.d.f.’s of Z, U , V , and W are, respectively, given by
+∞

80
Z
fZ (z) = f (x, z − x)dx,
−∞
Z +∞
fU (u) = f (u + y, y)dy,
−∞
Z +∞
v 1
fV (v) = f (x, ) dx,
x |x|
81
fW (w) =

3. Theorem : (Sum of Gamma random variables)

Z
−∞
+∞

−∞
f (xw, x)|x|dx.

Let X1 , X2 , . . . , Xr be independent random variables with Xi ∼ Gamma(αi , λ), ∀1 ≤ i ≤ r ;

then X1 + X2 + . . . + Xr ∼ Gamma(α1 + α2 + . . . + αr , λ).
0
Corollary 1 :
Let X1 , X2 , . . . , Xr be independent random variables with Xi ∼ exp(λ), ∀1 ≤ i ≤ r ;
then X1 + X2 + . . . + Xr ∼ Gamma(r , λ).
79

Corollary 2 :
Let X1 , X2 , . . . , Xr be independent random variables with Xi ∼ χ2 (ni ), ∀1 ≤ i ≤ r ;
then X1 + X2 + . . . + Xr ∼ χ2 (n1 + n2 + . . . + nr ).
4. Theorem : (Sum of Cauchy random variables)
Let X1 , X2 , . . . , Xr be independent random variables with Xi ∼ Cauchy (µi , θi ), ∀1 ≤ i ≤ r ;
then X1 + X2 + . . . + Xr ∼ Cauchy (µ1 + µ2 + . . . + µr , θ1 + θ2 + . . . + θr ).
i(

5. Theorem : (Sum of Normal random variables)

Let X1 , X2 , . . . , Xr be independent random variables with Xi ∼ N(µi , σi2 ), ∀1 ≤ i ≤ r ;
then X1 + X2 + . . . + Xr ∼ N (µ1 + µ2 + . . . + µr , σ12 + σ22 + . . . + σr2 ).
Corollary 1 :
Let X1 , X2 , . . . , Xr be independent random variables with Xi ∼ N (µ , σ 2 ), ∀1 ≤ i ≤ r ;
X1 + X2 + . . . + Xr σ2
then X1 + X2 + . . . + Xr ∼ N (rµ , rσ 2 ) and ∼ N (µ , ).
pa

r r
Corollary 2 :
Let X1 , X2 , . . . , Xr be independent random variables with Xi ∼ N (0,1), ∀1 ≤ i ≤ r ;
X1 + X2 + . . . + Xr
then √ ∼ N (0,1).
r
6. Theorem :
Let X ∼ N (0, 1), then X 2 ∼ χ2 (1) .
Pa

7. Theorem :
Let X and Y be iid N (0, σ 2 ) random variables. Then X/Y ∼ Cauchy (1, 0).

CAREER FIELD 8
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
8. Theorem :
Let X ∼ Cauchy (µ , 0), then λ/X ∼ Cauchy ( |λ|
µ , 0) , where λ is a constant.
Corollary 1 :
X ∼ Cauchy (1, 0) iff 1/X ∼ Cauchy (1, 0) .
9. Theorem :

66
Let X1 and X2 be two independent continuous random variables where X1 ∼ χ2 (k1 ) and X2 ∼ χ2 (k2 ).
X1 /k1
Then ∼ F(k1 , k2 ) .
X2 /k2
10. Theorem :
Let X and Y be two independent continuous random variables where X ∼ N (0, 1) and Y ∼ χ2 (k). Then
X
∼ Student-t (k) .

80
p
Y /k
11. Theorem :
Let X and Y be two independent continuous random variables where X ∼ exp (λ) and Y ∼ exp (λ). Then
X
∼ U(0, 1) .
X +Y
12. Theorem :

Gamma (α2 , λ). Then

X
X +Y
81
Let X and Y be two independent continuous random variables where X ∼ Gamma (α1 , λ) and Y ∼
∼ Beta (α1 , α2 ) .

13. Theorem : Let X and Y be two random variables with joint p.d.f. f (x, y) and E(X) , E(Y ) exists.
(i) For any c ∈ R, then E(c) = c .
(ii) For any c ∈ R, E(cX) exists and E(cX) = cE(X).
(iii) E(X +Y ) exists and E(X +Y ) = E(X)+E(Y ). [from (i),(ii),(iii) we can say E(aX +b) = aE(X)+b]
0
(iv) E(|X|) ≥ |E(X)|.
14. Theorem : If X, Y , Z, Xi ’s are random variables, then
79
(i) E(aX + b) = aE(X) + b, where a, b ∈ R,
(ii) V ar(aX + b) = a2 V ar(X), where a, b ∈ R,
(iii) Cov(aX + b, cY + d) = acCov(X, Y ), where a, b, c, d ∈ R,
(iv) V ar(X + Y ) = V ar(X) + V ar(Y ) + 2Cov(X, Y ),
(v) Cov(X,PY + Z) = P Cov(X, Y ) + Cov(X, Z), P
n n P
(vi) V ar( i=1 Xi ) = 1≤i,j≤n Cov(Xi , Xj ) = i=1 V ar(Xi ) + 2 1≤i<j≤n Cov(Xi , Xj ) =
Pn Pn−1 Pn
i=1 V ar(Xi ) + 2 i=1 j=i+1 Cov(Xi , Xj ),
i(

ab
(vii) ρaX,bY = |ab| ρX,Y , where a, b ∈ R.

15. Theorem : If X and Y are two independent random variable, then

(i) E(XY ) = E(X)E(Y ),
(ii) Cov(X, Y ) = 0,
(iii) ρX,Y = 0,
(iv) V ar(X + Y ) = V ar(X) + V ar(Y ).
pa

16. Theorem : (Cauchy-Schwarz Inequality) p

If X and Y be two random variables, then |E(XY )| ≤ E(X 2 )E(Y 2 ),
equality holds iff X = aY for some constant a ∈ R.
17. Theorem : If X and Y be two random variables, then |ρX,Y | ≤ 1,
equality holds iff X = aY for some constant a ∈ R.

18. Theorem : (Markov’s Inequality)

Pa

Let X be a nonnegative random variable having finite expectation, and let t be a positive real number.
E(X)
Then P (X ≥ t) ≤ .
t

CAREER FIELD 9
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
19. Theorem : (Chebyshev’s Inequality)
Let X be a random variable with expectation µ and variance σ 2 , let t be a positive real number. Then
σ2
P (|X − µ| ≥ t) ≤ 2 .
t
20. Theorem : (Weak Law of Large Numbers or WLLN)

66
Let X1 , X2 , . . . , Xn are i.i.d. random variables with common mean µ. For all ϵ > 0, P (|X¯n − µ| ≥ ϵ) → 0
as n → ∞ or equivalently, P (|X¯n − µ| < ϵ) → 1 as n → ∞.
21. Theorem : (Strong Law of Large Numbers or SLLN)
Let X1 , X2 , . . . , Xn are i.i.d. random variables with common mean µ. The sample mean X¯n converges to
µ pointwise as n → ∞, with probability 1.
22. Theorem : (Central Limit Theorem or CLT)

80
Let X1 , X2 , ... Xn be a sequence of i.i.d. random variables, each having mean µ and variance!σ 2 .
√ X¯n − µ
 
X1 + X2 + ... + Xn − nµ
Then the distribution of the random variable √ = n tends to
σ n σ
N (0, 1) as n → ∞. That is for a ∈ R,
  Z a
X1 + X2 + ... + Xn − nµ 1 2
lim P √ ≤ a = Φ(a), where Φ(a) = √ e−x /2 dx.
σ n

3
n→∞

PROBLEMS ———————————————————————
81
1. Let X follows N (0, 1). Find E(X k ), where k is an odd number.
2π −∞

2. Let a point be chosen randomly from the interior of a disk of radius R in the plane. Let X denote the
square of the distance of the point chosen from the center of the disk. Find the distribution function
0
(c.d.f.) of X.
3. Let a point be chosen uniformly from a solid ball in three-dimensional space of radius R. Let X denote
the distance of the point chosen from the clenter of the ball. Find the distribution function of X.
79

4. Let a point be chosen uniformly over the interval [0, a]. Let X denote the distance of the point chosen
from the origin. Find the distribution function of X. Find the distribution function of Y = M in(X, a/2).
5. Let a point be chosen uniformly from the interior of a triangle having a base of length l and height h
from the base. Let X be defined as the distance from the point chosen to the base. Find the [Link]
function of X.
i(

6. Consider an equilateral triangle whose sides each have length s. Let a point be chosen uniformly from one
side of the triangle. Let X denote the distance of the point chosen from the opposite vertex. Find the
distribution function of X.

7. Let X denote the sine of an angle chosen at random from (−π/2, π/2). Find the c.d.f. and p.d.f. function
of X.
pa

8. Let X be a continuous random variable having symmetric p.d.f. f (means f (x) = f (−x)) and such that
X 2 has an exponential density with parameter λ. Find f .

9. Let X follows N (0, σ 2 ). Find the density (p.d.f.) of Y = |X|.

10. Let X follows N (µ, σ 2 ). Find the density of Y = eX . [Here we say, this X follows Lognormal distribution].
11. Suppose that the weight of a person selected at random from some population is normally distributed
with parameters µ and σ. Suppose also that P (X ≤ 160) = 1/2 and P (X ≤ 140) = 1/4. Find µ and σ
Pa

and find P (X ≥ 200). Of all the people in the population weighing at least 200 pounds, what percentage
will weigh over 220 pounds?
12. Let X follows the Gamma distribution Γ(α, λ). Find the density of Y = cX, where c > O.

CAREER FIELD 10
Papai’s Academy of Data Science
 Papai (Bmath, Mmath, Phd Scholar Computer Science: ISI Institute)
Classes for ISI / CMI entrances Phn No- 7908180669, Email id- [Link]@[Link]

9)
√
13. Let X follows the Gamma distribution Γ(α, λ). Find the density of Y = X .
14. Let X follows the Gamma distribution Γ(α, λ), α > 0, λ > 0. Show that the Gamma p.d.f. has a
maximum value at (α − 1)/λ.
15. Let X and Y have the joint p.d.f. f given by

66
2
−xy+y 2 )/2
f (x, y) = ke−(x , − ∞ < x, y < ∞.

(i) Find the value of k.

(ii) Find the marginal p.d.f.’s of X and Y .
(iii) Prove that X and Y are not independent random variables.
(iv) Find the conditional distributions of Y /X=x (Y given X = x ) .

80
(v) Find E(X), E(Y ), V ar(X), V ar(Y ), E(XY ) and ρX,Y .
16. Let X be a uniformly distributed random variable over (0, 1), and let Y be a uniformly distributed random
variable over (0, X). Find the joint distribution of X , Y and the marginal distribution of Y .
17. Suppose X follows Logistic Distribution, which has c.d.f.
ex
x ∈ R.

(i) Find the p.d.f. of X.

(ii) Find P (−2 < X < 2).
F (x) =
81 1 + ex

18. Suppose X follows Rayleigh Distribution, which has c.d.f.

,

2
F (x) = 1 − e−x ,
0
x > 0.

(i) Find the p.d.f. of X.

(ii) Find P (X > 2).
79

(iii) Find E(X).

19. Two candidates A and B are fighting an election, where it is known that 55% of the electorate favor
candidate B. What is the probability that in a sample of size 100 at least one-half of those sampled will
favor candidate A ? [Use Central Limit Theorem and Standard Normal distribution table]
20. Let U1 and U2 be independent random variables each follows exponential distribution with parameter λ,
i(

and set Y = max(U1 , U2 ). Find the mean and variance of Y .

21. According to the data of February 11, 2020, the ages of COVID-19 patients in China are roughly normally
distributed with a mean of 51 and a standard deviation of 16.5.
(1) What is the probability that a patient has age between 40 to 49 years.
(2) What is the probability that a patient has age above 80 years.
(3) There were 44, 672 confirmed cases of the novel coronavirus COVID-19 in China at that time. Find
pa

the approximate value of number of patients with ages between 20 to 25 years.

[Use Standard Normal distribution table]
22. The average number of acres burned by forest and range fires in a large New Mexico county is 4,300
acres per year, with a standard deviation of 750 acres. The distribution of the number of acres burned is
normal. Given that, Φ(0.13) = 0.55, Φ(2.4) = 0.99 and Φ(0.31) = 0.62.
(i) What is the probability that between 2,500 and 4,200 acres will be burned in any given year?
Pa

(ii) What is number of burnt acres corresponds to the 38th percentile?

23. A die is rolled 420 times. What is the probability that the sum of the rolls lies between 1400 and 1550?
[Use Central Limit Theorem and Standard Normal distribution table]

CAREER FIELD 11
Papai’s Academy of Data Science